BGP Commands

This chapter describes the commands used to configure and monitor Border Gateway Protocol (BGP) on Cisco ASR 9000 Series Aggregation Services Routers using Cisco IOS XR software. The commands in this module configure IP Version 4 (IPv4), IP Version 6 (IPv6), Virtual Private Network Version 4 (VPNv4) routing sessions.

For detailed information about BGP concepts, configuration tasks, and examples, see the Implementing BGP chapter in the Routing Configuration Guide for Cisco ASR 9000 Series Routers.


Note


Running the show bgp command immediately after configuring a large and complex route policy may result in timeout of the system database shown through an error message (SYSDB-SYSDB-6-TIMEOUT_EDM). It is recommended that the show command be run after the new route policy takes effect.


accept-own

To enable handling of self-originated VPN routes containing ACCEPT_OWN community attribute, use the accept-own command in neighbor VPNv4 or VPNv6 address family configuration mode. To disable this functionality, either use the no form of this command or use the command with inheritance-disable keyword.

accept-own [inheritance-disable]

no accept-own

Syntax Description

inheritance-disable

Disables handling of self-originated VPN routes containing ACCEPT_OWN community attribute and prevents inheritance of Accept Own from a parent configuration.

Command Default

Disabled

Command Modes

Neighbor address family VPNv4

Neighbor address family VPNv6

Command History

Release Modification

Release 4.1.1

This command was introduced.

Usage Guidelines

To use this command, you must be in a user group associated with a task group that includes appropriate task IDs. If the user group assignment is preventing you from using a command, contact your AAA administrator for assistance.

Task ID

Task ID Operation

bgp

read, write

Examples

This example shows how to enable handling of accept-own community:


RP/0/RSP0/CPU0:router#configure
RP/0/RSP0/CPU0:router(config)#router bgp 100
RP/0/RSP0/CPU0:router(config-bgp)#neighbor 10.2.3.4
RP/0/RSP0/CPU0:router(config-bgp-nbr)#address-family vpnv4 unicast
RP/0/RSP0/CPU0:router(config-bgp-nbr-af)#accept-own

additional-paths install backup

To install a backup path into the forwarding table and provide prefix independent convergence (PIC) in case of a PE-CE link failure, use the additional-paths install backup command in an appropriate address family configuration mode. To prevent installing the backup path, use the no form of this command. To disable prefix independent convergence, use the disable keyword.

additional-paths install backup [disable]

no additional-paths install backup

Syntax Description

disable

Disables installing backup path into the forwarding table.

Command Default

None

Command Modes

VRF IPv4 address family configuration

VRF IPv6 address family configuration

VPNv4 address family configuration

VPNv6 address family configuration

Command History

Release

Modification

Release 3.9.0

This command was introduced.

Usage Guidelines

To use this command, you must be in a user group associated with a task group that includes appropriate task IDs. If the user group assignment is preventing you from using a command, contact your AAA administrator for assistance.

Task ID

Task ID

Operations

bgp

read, write

Examples

The following example shows how to enable installing a backup path into the forwarding table in VPNv4 address family mode:


RP/0/RSP0/CPU0:router#configure
RP/0/RSP0/CPU0:router(config)#router bgp 100
RP/0/RSP0/CPU0:router(config-bgp)#address-family vpnv4 unicast
RP/0/RSP0/CPU0:router(config-bgp-af)#additional-paths install backup

additional-paths receive

To configure receive capability of multiple paths for a prefix to the capable peers, use the additional-paths receive command in address-family configuration mode. To disable receive capability, use the no form of this command. To disable add-path receive capability for all neighbors belonging to a particular VRF address-family, use the disable option.

additional-paths receive [disable]

no additional-paths receive

Syntax Description

disable

Disables advertising additional paths receive capability.

Note

 

Use the disable keyword option to disable add-path receive capability for all neighbors belonging to a specified VRF address-family.

Syntax Description

This command has no keywords or arguments.

Command Default

None

Command Modes

IPv4 address family configuration

IPv6 address family configuration

VPNv4 address family configuration

VPNv6 address family configuration

VRF IPv4 address family configuration

VRF IPv6 address family configuration

Command History

Release Modification
Release 4.0.0

This command was introduced.

Usage Guidelines

To use this command, you must be in a user group associated with a task group that includes appropriate task IDs. If the user group assignment is preventing you from using a command, contact your AAA administrator for assistance.

Use the additional-paths receive command to allow add-path receive capability to be negotiated for a specified address family. When the additional-paths receive command is configured, the receive capability is automatically enabled for all internal BGP neighbors for a specified address family. When this command is either not configured or explicitly disabled, none of the neighbors are allowed to negotiate receive capability for the address family.

After enabling the receive capability, the session needs to be reset for the configuration to take into effect.

Task ID

Task ID Operation

bgp

read, write

Examples

This example shows how to enable additional paths receive capability under VPNv4 unicast address family:


RP/0/RSP0/CPU0:router# configure
RP/0/RSP0/CPU0:routerconfig)# router bgp 100
RP/0/RSP0/CPU0:router(config-bgp)# address-family vpnv4 unicast
RP/0/RSP0/CPU0:router(config-bgp-af)# additional-paths receive

This example shows how to disable additional paths receive capability for all neighbors belonging to a particular VRF address-family (vrf1):

RP/0/RSP0/CPU0:router#configure
RP/0/RSP0/CPU0:router(config-bgp)#vrf vrf1
RP/0/RSP0/CPU0:router(config-bgp-vrf)#address-family ipv4 unicast
RP/0/RSP0/CPU0:router(config-bgp-vrf-af)#additional-paths receive disable

additional-paths selection

To configure additional paths selection mode for a prefix, use the additional-paths selection command in address-family configuration mode. To disable the additional-paths selection mode for a prefix, use the no form of this command. To disable the additional-paths selection mode for a particular VRF address-family, use the disable option.

additional-paths selection {route-policy route-policy-name | disable}

no additional-paths selection route-policy route-policy-name

Syntax Description

route-policy route-policy-name

Specifies the name of a route policy used for additional paths selection.

disable

Disables add-path selection for a particular VRF address-family.

Command Default

None

Command Modes

IPv4 address family configuration

IPv6 address family configuration

VPNv4 address family configuration

VPNv6 address family configuration

VRF IPv4 address family configuration

VRF IPv6 address family configuration

Command History

Release Modification
Release 4.0.0

This command was introduced.

Usage Guidelines

To use this command, you must be in a user group associated with a task group that includes appropriate task IDs. If the user group assignment is preventing you from using a command, contact your AAA administrator for assistance.

To configure additional paths selection mode for some or all prefixes, use the additional-paths selection command by specifying a route-policy.

Use the additional-path selection command with an appropriate route-policy to calculate backup paths and to enable Prefix Independent Convergence (PIC) functionality. Refer BGP Prefix Independent Convergence Unipath Primary/Backup section in Routing Configuration Guide for Cisco ASR 9000 Series Routers for details on the PIC functionality.

Task ID

Task ID Operation

bgp

read, write

Examples

This example shows how to enable selection of additional paths:


RP/0/RSP0/CPU0:router# configure
RP/0/RSP0/CPU0:router(config)# router bgp 100
RP/0/RSP0/CPU0:router(config-bgp)# address-family vpnv4 unicast
RP/0/RSP0/CPU0:router(config-bgp-af)# additional-paths selection route-policy ap1 

This example shows how to disable add-path selection for a particular VRF address-family (vrf1):
RP/0/RSP0/CPU0:router#configure
RP/0/RSP0/CPU0:router(config-bgp)#vrf vrf1
RP/0/RSP0/CPU0:router(config-bgp-vrf)#address-family ipv4 unicast
RP/0/RSP0/CPU0:router(config-bgp-vrf-af)#additional-paths selection disable
This example shows how to enable add-path selection for a particular VRF address-family (vrf2):
RP/0/RSP0/CPU0:router#configure
RP/0/RSP0/CPU0:router(config-bgp)#vrf vrf2
RP/0/RSP0/CPU0:router(config-bgp-vrf)#address-family ipv4 unicast
RP/0/RSP0/CPU0:router(config-bgp-vrf-af)#additional-paths selection route-policy ap2

additional-paths send

To configure send capability of multiple paths for a prefix to the capable peers, use the additional-paths send command in address-family configuration mode. To disable the send capability, use the no form of this command.

additional-paths send [disable]

no additional-paths send

Syntax Description

disable

Disables advertising additional paths send capability.

Note

 

Use the disable option to disable add-path send capability for all neighbors belonging to a particular VRF address-family.

Command Default

None

Command Modes

IPv4 address family configuration

IPv6 address family configuration

VPNv4 address family configuration

VPNv6 address family configuration

VRF IPv4 address family configuration

VRF IPv6 address family configuration

Command History

Release Modification
Release 4.0.0

This command was introduced.

Usage Guidelines

To use this command, you must be in a user group associated with a task group that includes appropriate task IDs. If the user group assignment is preventing you from using a command, contact your AAA administrator for assistance.

Use the additional-paths send command to allow add-path send capability to be negotiated for a specified address family. When the additional-paths send command is configured, the send capability is automatically enabled for all internal BGP neighbors for the specified address family. When the command is either not configured or explicitly disabled, none of the neighbors are allowed to negotiate send capability for the address family.

After enabling the send capability, the session needs to be reset for the configuration to take into effect.

Task ID

Task ID Operation

bgp

read, write

Examples

This example shows how to enable additional paths send capability under VPNv4 4 unicast address family:


RP/0/RSP0/CPU0:router# configure
RP/0/RSP0/CPU0:routerconfig)# router bgp 100
RP/0/RSP0/CPU0:router(config-bgp)# address-family vpnv4 unicast
RP/0/RSP0/CPU0:router(config-bgp-af)# additional-paths send

This example shows how to enable add-path selection for a particular VRF address-family (vrf1):
RP/0/RSP0/CPU0:router#configure
RP/0/RSP0/CPU0:router(config-bgp)#vrf vrf1
RP/0/RSP0/CPU0:router(config-bgp-vrf)#address-family ipv4 unicast
RP/0/RSP0/CPU0:router(config-bgp-vrf-af)#additional-paths send disable

address-family (BGP)

To enter various address family configuration modes while configuring Border Gateway Protocol (BGP), use the address-family command in an appropriate configuration mode. To disable support for an address family, use the no form of this command.

address-family {ipv4 {labeled-unicast | multicast | unicast} | ipv6 {unicast} | l2vpn vpls-vpws | vpnv4 { unicast} | vpnv6 {unicast | }}

no address-family

Syntax Description

ipv4 unicast

Specifies IP Version 4 (IPv4) unicast address prefixes.

ipv4 multicast

Specifies IPv4 multicast address prefixes.

ipv4 labeled-unicast

Specifies IPv4 labeled-unicast address prefixes. This option is available in IPv4 neighbor configuration mode and VRF neighbor configuration mode.

ipv6 unicast

Specifies IP Version 6 (IPv6) unicast address prefixes.

vpnv4 unicast

Specifies VPN Version 4 (VPNv4) unicast address prefixes. This option is not available in VRF or VRF neighbor configuration mode.

l2vpn vpls-vpws

Specifies L2VPN vpls-vpws address prefixes.

Command Default

An address family must be explicitly configured in the router configuration mode for the address family to be active in BGP. Similarly, an address family must be configured under the neighbor for the BGP session to be established for that address family. An address family must be configured in router configuration mode before it can be configured under a neighbor.

Command Modes

Router configuration

Neighbor configuration

Neighbor group configuration

VRF configuration

VRF neighbor configuration (IPv4 address families)

Command History

Release

Modification

Release 3.7.2

This command was introduced.

Release 3.9.0

L2VPN Address Family support was added.

Usage Guidelines

To use this command, you must be in a user group associated with a task group that includes appropriate task IDs. If the user group assignment is preventing you from using a command, contact your AAA administrator for assistance.

Use the address-family command to enter various address family configuration modes while configuring BGP routing sessions. When you enter the address-family command from router configuration mode, you enable the address family and enter global address family configuration mode.

The IPv4 unicast address family must be configured in router configuration mode before configuring the IPv4 labeled-unicast address family for a neighbor in neighbor configuration mode.

Table 1. Address Family Submode Support

Address Family

Supported in Router Submode

Supported in Neighbor Submode

Comments

ipv4 unicast

yes

yes

ipv4 multicast

yes

yes

ipv4 labeled-unicast

no

yes

The ipv4 labeled-unicast address family can be configured only as a neighbor address family; however, it requires that the ipv4 unicast address family be configured as the router address family first.

vpnv4 unicast

yes

yes

ipv6 unicast

yes

yes

ipv6 multicast

yes

yes

l2vpn vpls-vpws

yes

yes

When you enter the address-family command from neighbor configuration mode, you activate the address family on the neighbor and enter neighbor address family configuration mode. IPv4 neighbor sessions support IPv4 unicast, , labeled-unicast, and VPNv4 unicast address families. IPv6 neighbor sessions support IPv6 unicast address families.

Outbound Route Filter (ORF) capability is not supported with address-family l2vpn vpls-vpws

Task ID

Task ID

Operations

bgp

read, write

Examples

The following example shows how to place the router in global address family configuration mode for the IPv4 address family:


Router# configure
Router(config)# router bgp 100
Router(config-bgp)# address-family ipv4 unicast
Router(config-bgp-af)#  

The following example shows how to activate IPv4 multicast for neighbor 10.0.0.1 and place the router in neighbor address family configuration mode for the IPv4 multicast address family:

Router# configure
Router# router bgp 1
Router(config-bgp)# address-family ipv4 multicast
Router(config-bgp-af)# exit 
Router(config-bgp)# neighbor 10.0.0.1
Router(config-bgp-nbr)# remote-as 1
Router(config-bgp-nbr)# address-family ipv4 multicast
Router(config-bgp-nbr-af)# 
  
The following example shows how to place the router in global address family configuration mode for the IPv4 tunnel address family:

Router# configure
Router(config)# router bgp 12
Router(config-bgp)# address-family ipv4 tunnel
Router(config-bgp-af)#  

advertise

To configure advertisement of local or re-originated VPNv4 or VPNv6 unicast routes or disable advertisement of L2VPN prefixes from a BGP router to its configured BGP neighbor, use the advertise command in BGP neighbor address family configuration mode. To undo this command configuration, use the no form of this command.

advertise { {vpnv4 | vpnv6} unicast {re-originated | local stitching-rt} | l2vpn evpn disable}

no advertise { {vpnv4 | vpnv6} unicast {re-originated | local stitching-rt} | l2vpn evpn disable}

Syntax Description

vpnv4

Specifies VPNv4 prefixes.

vpnv6

Specifies VPNv6 prefixes.

unicast

Specifies VPNv4 or VPNv6 unicast routes.

re-originated

Specifies advertisement of re-originated VPNv4 or VPNv6 unicast routes

local stitching-rt

Specifies advertisement of local VPNv4 or VPNv6 unicast routes with stitching route target identifier.

l2vpn

Specifies L2VPN address-family.

evpn disable

Disables advertisement of L2VPN EVPN prefixes.

Command Default

None

Command Modes

BGP neighbor address family configuration mode

Command History

Release Modification

Release 5.3.2

This command was introduced.

Usage Guidelines

To use this command, you must be in a user group associated with a task group that includes appropriate task IDs. If the user group assignment is preventing you from using a command, contact your AAA administrator for assistance.

Task ID

Task ID Operation

bgp

read, write

Examples

The following example shows how to configure a BGP router to advertise local VPNv4 unicast routes assigned with stitching route target identifier to the specified BGP neighbor 1.1.1.1.


RP/0/RSP0/CPU0:router# configure
RP/0/RSP0/CPU0:router(config)# router bgp 1
RP/0/RSP0/CPU0:router(config-bgp)# neighbor 1.1.1.1
RP/0/RSP0/CPU0:router(config-bgp-nbr)# address-family l2vpn evpn 
RP/0/RSP0/CPU0:router(config-bgp-nbr-af)# advertise vpnv4 unicast re-originated stitching-rt

advertise best-external

To advertise the best–external path to the iBGP and route-reflector peers, when a locally selected bestpath is from an internal peer, use the advertise best-external command in an appropriate address family configuration mode. To prevent advertising the best–external path, use the no form of this command. To disable advertising the best–external path, use the disable keyword.

advertise best-external [disable]

no advertise best-external

Syntax Description

disable

Disables best–external configuration for the VRF.

Command Default

None

Command Modes

VRF IPv4 address family configuration

VRF IPv6 address family configuration

L2VPN address family configuration

VPNv4 address family configuration

VPNv6 address family configuration

IPv4 labelled unicast configuration

IPv6 labelled unicast configuration

Command History

Release

Modification

Release 3.9.0

This command was introduced.

Usage Guidelines

To use this command, you must be in a user group associated with a task group that includes appropriate task IDs. If the user group assignment is preventing you from using a command, contact your AAA administrator for assistance.

Unlabelled best-external is not supported as it may create routing loop.

Task ID

Task ID

Operations

bgp

read, write

Examples

The following example shows how to enable advertising the best–external path VPNv4 unicast address family mode:


RP/0/RSP0/CPU0:routerconfigure
RP/0/RSP0/CPU0:router(config)# router bgp 100
RP/0/RSP0/CPU0:router(config-bgp)# address-family vpnv4 unicast
RP/0/RSP0/CPU0:router(config-bgp-af)# advertise best-external
  
 

advertise permanent-network

To identify the peers to whom the permanent paths must be advertised, use the advertise permanent-network command in the neighbor address family configuration mode. To stop advertising the permanent p, use the no form of this command. The permanent paths will always be advertised to peers having advertise permanent-network configuration, even if a different best-path is available. The permanent path is not advertised to peers that are not configured to receive permanent path.

The permanent path supports only prefixes in IPv4 unicast and IPv6 unicast address-families under the default Virtual Routing and Forwarding (VRF).

advertise permanent-network

no advertise permanent-network

Syntax Description

This command has no arguments or keywords.

Command Modes

Neighbor address-family configuration.

Command History

Release

Modification

Release 5.1.1

This command was introduced.

Usage Guidelines

To use this command, you must be in a user group associated with a task group that includes appropriate task IDs. If the user group assignment is preventing you from using a command, contact your AAA administrator for assistance.

Task ID

Task ID

Operations

bgp

read, write

Examples

This example shows how to advertise permanent path:


RP/0/RSP0/CPU0:router# configure
RP/0/RSP0/CPU0:router(config)# router bgp 100
RP/0/RSP0/CPU0:router(config-bgp)# neighbor 10.1.1.1
RP/0/RSP0/CPU0:router(config-bgp-nbr)# remote-as 4713
RP/0/RSP0/CPU0:router(config-bgp-nbr)# address-family ipv4 unicast
RP/0/RSP0/CPU0:router(config-bgp-nbr-af)# advertise permanent-network


advertise local-labeled-route safi-unicast disable

To disable the advertisement of a route with a local label to peers on the unicast subaddress Family Identifier (SAFI), use the advertise local-labeled-route safi-unicast disable command in an appropriate address family configuration mode.

advertise local-labeled-route safi-unicast disable

Syntax Description

This command has no arguments or keywords.

Command Default

The default is to allow the path with local label to be advertised on the unicast SAFI.

Command Modes

IPv4 address family configuration

IPv6 address family configuration

Command History

Release Modification

Release 6.2.1

This command was introduced.

Usage Guidelines

To use this command, you must be in a user group associated with a task group that includes appropriate task IDs. If the user group assignment is preventing you from using a command, contact your AAA administrator for assistance.

Examples

To disable the advertisement of the route with a local label to peers on the unicast SAFI:


Router(config)# configure
Router(config)# router bgp 1
Router(config-bgp)# address-family ipv4 unicast 
Router(config-bgp-af)# advertise local-labeled-route safi-unicast disable
 

advertisement-interval

To set the minimum interval between the sending of Border Gateway Protocol (BGP) routing updates, use the advertisement-interval command in an appropriate configuration mode. To remove the advertisement-interval command from the configuration file and restore the system to its default interval values, use the no form of this command.

advertisement-interval seconds

no advertisement-interval [seconds]

Syntax Description

seconds

Minimum interval between sending BGP routing updates (in seconds). Range is 0 to 600.

Command Default

Default minimum interval:

For internal BGP (iBGP) peers is 0 seconds

For external BGP (eBGP) peers is 30 seconds

For customer edge (CE) peers is 0 seconds

Command Modes

Neighbor configuration

Neighbor group configuration

Session group configuration

VRF neighbor configuration

Command History

Release

Modification

Release 3.7.2

This command was introduced.

Release 5.0.0

This command was introduced.

Usage Guidelines

To use this command, you must be in a user group associated with a task group that includes appropriate task IDs. If the user group assignment is preventing you from using a command, contact your AAA administrator for assistance.

If this command configures a neighbor group or session group, all neighbors using the group inherit the configuration. Values of commands configured specifically for a neighbor override inherited values.

Task ID

Task ID

Operations

bgp

read, write

Examples

The following example shows how to set the minimum time between sending BGP routing updates to 10 seconds:


RP/0/RSP0/CPU0:router(config)# router bgp 5
RP/0/RSP0/CPU0:router(config-bgp)# neighbor 10.1.1.1 
RP/0/RSP0/CPU0:router(config-bgp-nbr)# remote-as 100
RP/0/RSP0/CPU0:router(config-bgp-nbr)# advertisement-interval 10

af-group

To create an address family group for Border Gateway Protocol (BGP) neighbors and enter address family group configuration mode, use the af-group command in router configuration mode. To remove an address family group, use the no form of this command.

af-group af-group-name address-family

no af-group

Syntax Description

af-group-name

Address family group name.

address-family

Enters address family configuration mode.

ipv4 unicast

Specifies IP Version 4 (IPv4) unicast address prefixes.

ipv4 multicast

Specifies IPv4 multicast address prefixes.

ipv4 labeled-unicast

Specifies IPv4 labeled unicast address prefixes.

ipv4 tunnel

Specifies IPv4 tunnel address prefixes.

ipv6 unicast

Specifies IP Version 6 (IPv6) unicast address prefixes.

vpnv4 unicast

Specifies VPN Version 4 (VPNv4) unicast address prefixes.

Command Default

No BGP address family group is configured.

Command Modes

Router configuration

Command History

Release

Modification

Release 3.7.2

This command was introduced.

Usage Guidelines

To use this command, you must be in a user group associated with a task group that includes appropriate task IDs. If the user group assignment is preventing you from using a command, contact your AAA administrator for assistance.

Use the af-group command to group address family-specific neighbor commands within an IPv4 or IPv6 address family. Neighbors that have address family configuration are able to use the address family group. Further, neighbors inherit the configuration parameters of the entire address family group.

You cannot define two address family groups with the same name in different address families.

Task ID

Task ID

Operations

bgp

read, write

Examples

The following example shows how to create address family group group1 and enter address family group configuration mode for IPv4 unicast. Group1 contains the next-hop-self feature, which is inherited by neighbors that use address family group1.


Router(config)# router bgp 100
Router(config-bgp)# af-group group1 address-family ipv4 unicast
Router(config-bgp-afgrp)# next-hop-self

aggregate-address

To create an aggregate entry in a Border Gateway Protocol (BGP) routing table, use the aggregate-address command in an appropriate configuration mode. To remove the aggregate-address command from the configuration file and restore the system to its default condition, use the no form of this command.

aggregate-address address/mask-length [as-set] [as-confed-set] [summary-only] [route-policy { route-policy-name | aggregate-route-policy-name } ]

no aggregate-address

Syntax Description

address

Aggregate address.

/mask-length

Aggregate address mask length.

as-set

(Optional) Generates autonomous system set path information and community information from contributing paths.

as-confed-set

(Optional) Generates autonomous system confederation set path information from contributing paths.

summary-only

(Optional) Filters all more-specific routes from updates.

route-policy route-policy-name

(Optional) Specifies the name of a route policy used to set the attributes of the aggregate route.

route-policy aggregate-route-policy-name

(Optional) Specifies the name of a aggregate route policy.

Command Default

When you do not specify this command, no aggregate entry is created in the BGP routing table.

Command Modes

IPv4 address family configuration

IPv6 address family configuration

VRF IPv4 address family configuration

VRF IPv6 address family configuration

Command History

Release

Modification

Release 3.7.2

This command was introduced.

Release 7.5.4

The command output displayed to set an aggregate contributor to a specific aggregate-address route.

Usage Guidelines

To use this command, you must be in a user group associated with a task group that includes appropriate task IDs. If the user group assignment is preventing you from using a command, contact your AAA administrator for assistance.

You can implement aggregate routing in BGP either by redistributing an aggregate route into BGP using the network command or the aggregate-address command.

Use the aggregate-address command without optional arguments to create an aggregate entry in the BGP routing table if any more-specific BGP routes are available that fall in the specified range. The aggregate route is advertised as coming from your autonomous system and has the atomic aggregate attribute set to show that information might be missing. (By default, the atomic aggregate attribute is set unless you specify the as-set keyword.)

Use of the as-set keyword creates an aggregate entry using the same rules that the command follows without this keyword. However, the advertised path for this route is an AS_SET, a set of all autonomous systems contained in all paths that are being summarized.

Do not use this form of the aggregate-address command when aggregating many paths because this route must be continually withdrawn and updated as autonomous system path reachability information for the summarized routes changes.

Use the as-confed-set keyword to create an AS_CONFED_SET in the autonomous system path of the aggregate from any confederation segments in the paths being summarized. This keyword takes effect only if the as-set keyword is also specified.

Use of the summary-only keyword creates an aggregate entry (for example, 10.0.0.0/8) but suppresses advertisements of more-specific routes to all neighbors. If you want to suppress only advertisements to certain neighbors, use the route-policy (BGP) command in neighbor address family configuration mode with caution. If a more-specific route leaks out, all BGP speakers (the local router) prefer that route over the less-specific aggregate you generate (using longest-match routing).

Use the route-policy keyword to specify a routing policy for the aggregate entry. The route-policy keyword is used to select which more-specific information to base the aggregate entry on and which more-specific information to suppress. You can also use the keyword to modify the attributes of the aggregate entry.

Aggregate contributor flag is set in the router where you configure it. This feature is applicable for the following Address Family Indicators (AFIs):

  • IPv4 unicast

  • IPv6 unicast

Task ID

Task ID

Operations

bgp

read, write

Examples

The following example shows how to create an aggregate address. The path advertised for this route is an autonomous system set consisting of all elements contained in all paths that are being summarized.


Router(config)# router bgp 100
Router(config-bgp)# address-family ipv4 unicast
Router(config-bgp-af)# aggregate-address 10.0.0.0/8 as-set

The following example shows to set an aggregate contributor to a specific aggregate-address route:

Router #config
Router(config)#router bgp 100
Router(config-bgp)#address-family ipv4 unicast
Router(config-bgp-af)#aggregate-address 250.2.2.0/24 route-policy aggregate-policy1
Router(config-bgp-af)#commit

aigp

To enable sending and receiving of accumulated interior gateway protocol (AiGP) attribute per eBGP neighbor, use the aigp command in appropriate configuration mode. To disable this functionality, either use the disable keyword or use the no form of this command.

aigp [disable]

no aigp

Syntax Description

disable

Disables sending or receiving AiGP attribute.

Command Default

Send or recive of AiGP attribute is disabled for eBGP neighbors

Command Modes

IPv4 address family configuration

IPv6 address family configuration

VRF IPv4 address family configuration

VRF IPv6 address family configuration

VPNv4 address family configuration

VPNv6 address family configuration

Neighbor address family configuration

VRF neighbor address family configuration

Command History

Release Modification
Release 4.0.0

This command was introduced.

Usage Guidelines

To use this command, you must be in a user group associated with a task group that includes appropriate task IDs. If the user group assignment is preventing you from using a command, contact your AAA administrator for assistance.

Task ID

Task ID Operation
bgp

read, write

Examples

The following example shows how to enable AiGP send and receive capability under neighbor address family (IPv4 unicast):


RP/0/RSP0/CPU0:routerconfigure
RP/0/RSP0/CPU0:router(config)# router bgp 100
RP/0/RSP0/CPU0:router(config-bgp)# neighbor 10.2.3.4
RP/0/RSP0/CPU0:router(config-bgp-nbr)# address-family ipv4 unicast 
RP/0/RSP0/CPU0:router(config-bgp-nbr-af)# aigp

aigp send-cost-community

To send Accumulated Interior Gateway Protocol (AiGP) value in cost community, use the aigp send-cost-community command in appropriate configuration mode. To disable sending AiGP value in cost community, either use the no form of this command or the disable keyword.

aigp send-cost-community {cost-id | disable} poi {igp-cost | pre-bestpath} [transitive]

no aigp send-cost-community

Syntax Description

cost-comm-id

Specifies the Cost community ID. The range is 0 to 255.

poi Point of insertion for bestpath calculation.
igp-cost Configures that cost community be used after iGP distance to next hop.
pre-bestpath Configures cost community as first step in best path calculation.
transitive

(Optional) Enables transitive cost community

disable

Disables sending AiGP value in cost community.

Command Default

Sending AiGP value in cost community is disabled

Command Modes

Neighbor address family configuration

VRF neighbor address family configuration

Command History

Release Modification
Release 4.0.0

This command was introduced.

Usage Guidelines

To use this command, you must be in a user group associated with a task group that includes appropriate task IDs. If the user group assignment is preventing you from using a command, contact your AAA administrator for assistance.

Cost community point of insertion can be configured either to be pre-bestpath or after igp cost. The transitive keyword is not required for iBGP sessions. However, the transitive keyword is required for eBGP sessions to convert AiGP metric into cost-community and advertise to the eBGP neighbors.

Task ID

Task ID Operation
bgp

read, write

Examples

The following example shows how to enable sending AiGP value in cost community ID 254 under neighbor address family (IPv4 unicast):


RP/0/RSP0/CPU0:routerconfigure
RP/0/RSP0/CPU0:router(config)# router bgp 100
RP/0/RSP0/CPU0:router(config-bgp)# neighbor 10.2.3.4
RP/0/RSP0/CPU0:router(config-bgp-nbr)# address-family ipv4 unicast 
RP/0/RSP0/CPU0:router(config-bgp-nbr-af)# aigp send-cost-community 254

allocate-label

To allocate Multiprotocol Label Switching (MPLS) labels for specific IPv4 unicast or IPv6 unicast or VPN routing and forwarding (VRF) IPv4 unicast routes so that the BGP router can send labels with BGP routes to a neighboring router configured for labeledunicast sessions, use the allocate-label command in the appropriate configuration mode. To restore the system to its default condition, use the no form of this command.

allocate-label {route-policy route-policy-name | all}

no allocate-label {route-policy route-policy-name | all}

Syntax Description

all

Allocates labels for all prefixes

route-policy route-policy-name

Uses a route policy to select prefixes for label allocation.

Command Default

No default behavior or values

Command Modes

IPv4 address family configuration

VRF IPv4 address family configuration

Command History

Release

Modification

Release 3.7.2

This command was introduced.

Usage Guidelines

To use this command, you must be in a user group associated with a task group that includes appropriate task IDs. If the user group assignment is preventing you from using a command, contact your AAA administrator for assistance.

Use the allocate-label command with a route policy to trigger BGP to allocate labels for all or a filtered set of global IPv4 routes (as dictated by the route policy). The command enables autonomous system border routers (ASBRs) that have labeled IPv4 unicast sessions to exchange Multiprotocol Label Switching (MPLS) labels with the IPv4 routes to the other autonomous system (AS) in Layer 3 Virtual Private Network (L3VPN) inter-AS deployments.


Note


The allocate-label all command is functionally equivalent to the allocate-label route-policy route-policy-name command when the route policy is a pass-all policy.


See MPLS Configuration Guide for Cisco ASR 9000 Series Routersfor information on using the allocate-label command for L3VPN inter-AS deployments and carrier-supporting-carrier IPv4 BGP label distribution.

Task ID

Task ID

Operations

bgp

read, write

Examples

The following example shows how to enable allocating labels for IPv4 routes:


RP/0/RSP0/CPU0:router(config)# router bgp 6
RP/0/RSP0/CPU0:router(config-bgp)# address family ipv4 unicast
RP/0/RSP0/CPU0:router(config-bgp-af)# allocate-label route-policy policy_A

allow vpn default-originate

To configure the router to be enabled to advertise a default route to a configured BGP VPN neighbor, use the allow vpn default-originate command in the BGP VRF Address-Family configuration mode. To undo this configuration, use the no form of this command.

allow vpn default-originate

no allow vpn default-originate

Syntax Description

This command has no keywords or arguments.

Command Default

The router cannot advertise a default route to its BGP VPN neighbors.

Command Modes

BGP VRF Address-Family configuration mode

Command History

Release Modification
Release 4.3.2

This command was introduced.

Usage Guidelines

This command only enables the router to advertise itself as the next-hop router for a default route to its BGP VPN neighbors. To actually forward the default route to a BGP VPN neighbor, you need to run the default-originate command under the BGP neighbor Address-Family configuration mode.

Task ID

Task ID Operation

bgp

read, write

Examples

The following example configuration shows how to enable a BGP router to advertise a default route to its BGP VPN neighbors.


RP/0/RSP0/CPU0:router# configure
RP/0/RSP0/CPU0:router(config)# router bgp 1
RP/0/RSP0/CPU0:router(config-bgp)# vrf foo
RP/0/RSP0/CPU0:router(config-bgp-vrf)# address-family ipv4 unicast
RP/0/RSP0/CPU0:router(config-bgp-vrf-af)# allow vpn default-originate

allowas-in

To allow an AS path with the provider edge (PE) autonomous system number (ASN) a specified number of times, use the allowas-in command in an appropriate configuration mode. To restore the system to its default condition, use the no form of this command.

allowas-in [as-occurrence-number]

no allowas-in [as-occurrence-number]

Syntax Description

as-occurrence-number

(Optional) Number of times a PE ASN is allowed. Range is 1 to 10.

Command Default

No default behavior or values

Command Modes

Address family group configuration

Neighbor address family configuration

Command History

Release

Modification

Release 3.7.2

This command was introduced.

Usage Guidelines

To use this command, you must be in a user group associated with a task group that includes appropriate task IDs. If the user group assignment is preventing you from using a command, contact your AAA administrator for assistance.

Hub and spoke VPN networks require looping back of routing information to the hub PE through the hub customer edge (CE). See MPLS Configuration Guide for Cisco ASR 9000 Series Routers for information on hub and spoke VPN networks. This looping back, in addition to the presence of the PE ASN, causes the looped-back information to be dropped by the hub PE.

The allowas-in command prevents the looped-back information from being dropped by replacing the neighbor autonomous system number (ASN) with the PE ASN in the AS path. This allows the VPN customer to see a specified number of occurrences of the PE ASN in the AS path.

Task ID

Task ID

Operations

bgp

read, write

Examples

The following example shows how to allow five occurrences of the PE ASN:


RP/0/RSP0/CPU0:router(config)# router bgp 6
RP/0/RSP0/CPU0:router(config-bgp)# af-group group_1 address-family vpnv4 unicast
RP/0/RSP0/CPU0:router(config-bgp-afgrp)# allowas-in 5

allowconfedas-in

To enable peer routers in the same confederation to learn from each other for a specified number of times, use the allowconfedas-in command in neighbor address family configuration mode.

allowconfedas-in confederation-as-count

Syntax Description

confederation-as-count

Number of times a provider edge (PE) confederation AS number is allowed.

Range: 1 to 10. Default: 3

Command Default

Peers within the same Confederation will not learn each other's routes.

Command Modes

Neighbor address-family configuration

Command History

Release Modification
Release 7.11.1

This command was introduced.

Task ID

Task ID Operation
bgp

read, write

Examples

In this example, the peer routers in the same confederation learn from each other thrice:

Router# router bgp 65001
Router(config-bgp)# bgp confederation peers 65002
Router(config-bgp)# bgp confederation identifier 100
Router(config-bgp)# neighbor 198.51.100.3
Router(config-bgp-nbr)# address-family ipv4 unicast
Router(config-bgp-nbr-af)# allowconfedas-in 1

as-format

To configure the router's Autonomous system number (ASN) notation to asdot format, use the as-format command in global configuration mode. To restore the system to its default condition, use the no form of this command.

as-format asdot

no

Syntax Description

asdot

Specifies the Autonomous system number (ASN) notation to asdot format.

Command Default

The default value, if the as-format command is not configured, is asplain.

Command Modes

Global configuration

Command History

Release

Modification

Release 3.9.0

This command was introduced.

Usage Guidelines

To use this command, you must be in a user group associated with a task group that includes appropriate task IDs. If the user group assignment is preventing you from using a command, contact your AAA administrator for assistance.

Task ID

Task ID

Operations

bgp

read, write

Examples

The following example shows how to configure the ASN notation to the asdot format:


RP/0/RSP0/CPU0:router(config)# as-format asdot

as-league peers

To configure a group of peer autonomous systems (AS) that will be used under common administration or a trusted relationship, use the as-league peers command in router configuration mode. To remove an AS from the group, use the no form of this command.

as-league peers as-number

Syntax Description

as-number

Autonomous system (AS) numbers for Border Gateway Protocol (BGP) peers that belong to this league.

  • Range for 2-byte ASNs is 1 to 65535.

  • Range for 4-byte ASNs in asplain format is 1 to 4294967295.

  • Range for 4-byte ASNs is asdot format is 1.0 to 65535.65535.

Command Default

None

Command Modes

Router configuration

Command History

Release Modification
Release 4.2.1

Introduced in release 4.2.1.

Usage Guidelines

Additional paths send and receive capabilities are negotiated per neighbor per address family. Originally, Add-path capability was supported only for iBGP and confederation neighbors and not supported for eBGP neighbors. Use the as-league peers command to configure a subset of peer ASs so that the eBGP peers belonging to that league can participate in the add path send/receive. You can create a subset of peer ASs that are under a common administration, have a trusted relationship, or enable a select set of features among themselves.

Task ID

Task ID Operation
bgp

read, write

Examples

This example shows how to create an AS league that includes a 200, 300, and 400. These are eBGP peers which, without an AS league peers configuration, do not negotiate additional paths send/receive. The following example shows how to create an AS league that includes peers 200, 300, and 400:

RP/0/RSP0/CPU0:router(config)# router bgp 100
RP/0/RSP0/CPU0:router(config-bgp)# as-league peers
RP/0/RSP0/CPU0:router(config-bgp-as-league-peers)# 200
RP/0/RSP0/CPU0:router(config-bgp-as-league-peers)# 300
RP/0/RSP0/CPU0:router(config-bgp-as-league-peers)# 400
RP/0/RSP0/CPU0:router(config-bgp-as-league-peers)#

 

!

as-override

To configure a provider edge (PE) router to override the autonomous system number (ASN) of a site with the ASN of a provider, use the as-override command which works for both VRF and non-VRF neighbor address family configuration mode. To restore the system to its default condition, use the no form of this command.

as-override [inheritance-disable]

no as-override [inheritance-disable]

Syntax Description

inheritance-disable

(Optional) Prevents the as-override command from being inherited from a parent group.

Command Default

Automatic override of the ASN is disabled.

Command Modes

VRF and non-VRF neighbor address family configuration

Command History

Release

Modification

Release 3.7.2

This command was introduced.

Release 3.9.0

The disable keyword was replaced with the inheritance-disable keyword.

Usage Guidelines

To use this command, you must be in a user group associated with a task group that includes appropriate task IDs. If the user group assignment is preventing you from using a command, contact your AAA administrator for assistance.

Use the as-override command in conjunction with the site-of-origin (SoO) feature, identifying the site where a route originated, and preventing routing loops between routers within a VPN.

Task ID

Task ID

Operations

bgp

read, write

Examples

The following example shows how to configure an ASN override:


RP/0/RSP0/CPU0:router(config)# router bgp 6
RP/0/RSP0/CPU0:router(config-bgp)# vrf vrf_A 
RP/0/RSP0/CPU0:router(config-bgp-vrf)# neighbor 192.168.70.24
RP/0/RSP0/CPU0:router(config-bgp-vrf-nbr)# remote-as 10 
RP/0/RSP0/CPU0:router(config-bgp-vrf-nbr)# address-family ipv4 unicast 
RP/0/RSP0/CPU0:router(config-bgp-vrf-nbr-af)# as-override 

 

as-path-loopcheck out disable

To disable AS PATH loop checking for outbound updates, use the as-path-loopcheck out disable command in an appropriate address family configuration mode. To re-enable the default AS PATH loop checking, use the no form of this command.

as-path-loopcheck out disable

no as-path-loopcheck out disable

Syntax Description

This command has no keywords or arguments.

Command Default

AS PATH loop checking for outbound updates is enabled if there is only one neighbor and disabled if there are multiple neighbors in the update group.

Command Modes

IPv4 address family

IPv6 address family

L2VPN address family

VPNv4 address family

VPNv6 address family

Command History

Release Modification

Release 3.8.2

This command was introduced.

Usage Guidelines

To use this command, you must be in a user group associated with a task group that includes appropriate task IDs. If the user group assignment is preventing you from using a command, contact your AAA administrator for assistance.

Configure the as-path-loopcheck out disable command to disable the default behavior of PE router not announcing BGP routes to the CE router if the routes contain an AS number matching the AS number of the receiving CE router.

Task ID

Task ID Operation

bgp

read, write

Examples

This example shows how to configure as-path-loopcheck out disable under IPv6 unicast address family:


RP/0/RSP0/CPU0:router#configure
RP/0/RSP0/CPU0:router(config)#router bgp 100
RP/0/RSP0/CPU0:router(config-bgp)#address-family ipv6 unicast 
RP/0/RSP0/CPU0:router(config-bgp-af)#as-path-loopcheck out disable

attribute-filter group

To configure attribute-filter group command mode, use the attribute-filter group command in an appropriate configuration mode. To disable attribute-filter group command mode, use the no form of this command.

attribute-filter group group-name

no attribute-filter group group-name

Syntax Description

group-name

Specifies the name of the attribute-filter group.

Command Default

Attribute-filter group command mode is disabled.

Command Modes

Router configuration

Neighbor configuration

Command History

Release Modification

Release 4.2.3

This command was introduced.

Usage Guidelines

To use this command, you must be in a user group associated with a task group that includes appropriate task IDs. If the user group assignment is preventing you from using a command, contact your AAA administrator for assistance.

Use the attribute-filter group command in neighbor configuration mode to configure a specific attribute filter group for a BGP neighbor.

Task ID

Task ID Operation

bgp

read, write

Examples

This example shows how to configure the attribute-filter group command mode:


Router#configure
Router(config)#router bgp 100
Router(config-bgp)#attribute-filter group ag_discard_med
Router(config-bgp-attrfg)#

This example shows how to configure the attribute filter group for a BGP neighbor:


Router#configure
Router(config)#router bgp 100
Router(config-bgp)#neighbor 10.0.1.101
Router(config-bgp-nbr)#remote-as 6461
Router(config-bgp-nbr)#update in filtering 
Router(config-nbr-upd-filter)#attribute-filter group ag_discard_med

attribute large-community

To specify the large community BGP path attributes in the created attribute filter and the action to be taken on the update message received from the BGP neighbor, use the attribute large-community command in the BGP attribute-filter group command mode.

attribute large-community { discard | treat-as-withdraw }

Syntax Description

discard

Specifies that the large-community BGP path attribute received in a BGP update message from the neighbor must be discarded.

treat-as-withdraw Indicates that the entire BGP update message that contains the large-community attribute has to be treated as a 'withdraw' message for the Network Layer Reachability Information (NLRI) contained in the update message.

Command Default

None

Command Modes

BGP attribute-filter group mode

Command History

Release

Modification

Release 6.3.1

This command was introduced.

Usage Guidelines

To use this command, you must be in a user group associated with a task group that includes appropriate task IDs. If the user group assignment is preventing you from using a command, contact your AAA administrator for assistance.

Task ID

Task ID Operation

bgp

read, write

Examples

The following example creates the attribute-filter named dogbert and specifies the large-community attribute and the action to discard. The dogbert attribute-filter is applied to the BGP neighbor 10.10.10.3. If the large-community BGP path attribute is received in a BGP update message from the neighbor 10.10.10.3, then the attribute will be discarded before further processing of the message.


RP/0/RSP0/CPU0:router#configure
RP/0/RSP0/CPU0:router(config)#router bgp 3
RP/0/RSP0/CPU0:router(config-bgp)#attribute-filter group dogbert
RP/0/RSP0/CPU0:router(config-bgp-attrfg)#attribute large-community discard 
RP/0/RSP0/CPU0:router(config-bgp-attrfg)#neighbor 10.10.10.3
RP/0/RSP0/CPU0:router(config-bgp-nbr)#update in filtering
RP/0/RSP0/CPU0:router(config-nbr-upd-filter)#attribute-filter group dogbert

bfd (BGP)

To specify a bidirectional forwarding detection (BFD) multiplier and minimum-interval arguments per neighbor, use the bfd command in neighbor address family independent configuration mode. To return to the system defaults, use the no form of this command.

Previous to this enhancement, BFD could be configured only in global scope in BGP. This change makes available two new command-line arguments under neighbor address family independent configuration:

bfd {multiplier | | minimum-interval} value

no bfd {multiplier | | minimum-interval} value

Syntax Description

multiplier value

Specifies the BFD session's multiplier value for the neighbor.

minimum-interval value

Specifies the BFD session's minimum-interval value for the neighbor.

Command Default

No default per neighbor parameters are set.

Command Modes

Neighbor address family independent configuration

Command History

Release

Modification

Release 3.7.2

This command was introduced.

Usage Guidelines

To use this command, you must be in a user group associated with a task group that includes appropriate task IDs. If the user group assignment is preventing you from using a command, contact your AAA administrator for assistance.

If the minimum interval is changed using the bfd minimum-interval command, the new parameter updates all affected BFD sessions under the command mode in which the minimum interval was changed.

If the multiplier is changed using the bfd multiplier command, the new parameter is used to update only the BFD sessions associated with the affected neighbor gets affected.

The assumption is that when BFD fast-detect is enabled under neighbor address family independent configuration, the values for the multiplier and minimum-interval values are always derived from the per-neighbor values if they are configured; otherwise, they are to be taken from the global BGP configuration mode. In the event that this has not been explicitly stated, then these values are taken to be the default values. Also, the bfd arguments can be configured under neighbor-group and session-group and the inheritance adheres to the standard way of BGP configuration inheritance.

Accordingly, there are four cases in which bfd-fast detect is enabled.

This is shown in table below where the BFD value is either multiplier or minimum-interval. Local indicates per NBR value, global is the BGP global value.

BFD value (global)

BFD value (local)

Result

Yes

Yes

BFD value (local)

Yes

No

BFD value (global)

No

Yes

BFD value (local)

No

No

BFD value (default)

Examples

The following example shows how to specify the BFD session's multiplier value for the neighbor:


RP/0/RSP0/CPU0:router# configure
RP/0/RSP0/CPU0:router(config)# router bgp 65000
RP/0/RSP0/CPU0:router(config-bgp-nbrgrp)#neighbor 3.3.3.2 
RP/0/RSP0/CPU0:router(config-bgp-nbr)# bfd minimum-interval 311
RP/0/RSP0/CPU0:router(config-bgp-nbr)# bfd multiplier 7
RP/0/RSP0/CPU0:router(config-bgp-nbr)# neighbor 5.5.5.2
RP/0/RSP0/CPU0:router(config-bgp-nbr)# bfd minimum-interval 318
RP/0/RSP0/CPU0:router(config-bgp-nbr)# bfd multiplier 4
RP/0/RSP0/CPU0:router(config-bgp-nbr)# vrf one
RP/0/RSP0/CPU0:router(config-bgp-vrf)# neighbor 3.12.1.2
RP/0/RSP0/CPU0:router(config-bgp-vrf-nbr)# bfd minimum-interval 119
RP/0/RSP0/CPU0:router(config-bgp-vrf-nbr)# bfd multiplier 10
RP/0/RSP0/CPU0:router(config-bgp-vrf-nbr)# commit
    
  RP/0/RSP0/CPU0:router# show bfd session
  Interface            Dest Addr           Local det time(int*mult)      State
                                             Echo            Async
  -------------------- --------------- ---------------- ---------------- ---------
  Gi0/2/0/2            3.3.3.2         2177ms(311ms*7)  14s(2s*7)        UP
  Gi0/2/0/2.1          3.12.1.2        1190ms(119ms*10) 20s(2s*10)       UP
  PO0/3/0/6            5.5.5.2         1272ms(318ms*4)  8s(2s*4)         UP
  
  
  RP/0/RSP0/CPU0:router# show bfd session detail
  I/f: GigabitEthernet0/2/0/2, Location: 0/2/CPU0, dest: 3.3.3.2, src: 3.3.3.1
   State: UP for 0d:0h:4m:44s, number of times UP: 1
  Received parameters:
   Version: 1, desired tx interval: 2 s, required rx interval: 2 s
   Required echo rx interval: 1 ms, multiplier: 7, diag: None
   My discr: 524295, your discr: 524296, state UP, D/F/P/C/A: 0/0/0/1/0
  Transmitted parameters:
   Version: 1, desired tx interval: 2 s, required rx interval: 2 s
   Required echo rx interval: 1 ms, multiplier: 7, diag: None
   My discr: 524296, your discr: 524295, state UP, D/F/P/C/A: 0/0/0/1/0
  Timer Values:
   Local negotiated async tx interval: 2 s
   Remote negotiated async tx interval: 2 s
   Desired echo tx interval: 311 ms, local negotiated echo tx interval: 311 ms
   Echo detection time: 2177 ms(311 ms*7), async detection time: 14 s(2 s*7)
  Local Stats:
   Intervals between async packets:
     Tx: Number of intervals=100, min=1664 ms, max=2001 ms, avg=1838 ms
         Last packet transmitted 313 ms ago
     Rx: Number of intervals=100, min=1662 ms, max=2 s, avg=1828 ms
         Last packet received 1615 ms ago
   Intervals between echo packets:
     Tx: Number of intervals=100, min=181 ms, max=462 ms, avg=229 ms
         Last packet transmitted 289 ms ago
     Rx: Number of intervals=100, min=178 ms, max=461 ms, avg=229 ms
         Last packet received 287 ms ago
   Latency of echo packets (time between tx and rx):
     Number of packets: 100, min=0 us, max=4 ms, avg=860 us
  Session owner information:
    Client           Desired interval         Multiplier
    ---------------- --------------------     --------------
    bgp-0            311 ms                   7
  
  I/f: GigabitEthernet0/2/0/2.1, Location: 0/2/CPU0, dest: 3.12.1.2, src: 3.12.1.1
   State: UP for 0d:0h:4m:44s, number of times UP: 1
  Received parameters:
   Version: 1, desired tx interval: 2 s, required rx interval: 2 s
   Required echo rx interval: 1 ms, multiplier: 10, diag: None
   My discr: 524296, your discr: 524295, state UP, D/F/P/C/A: 0/0/0/1/0
  Transmitted parameters:
   Version: 1, desired tx interval: 2 s, required rx interval: 2 s
   Required echo rx interval: 1 ms, multiplier: 10, diag: None
   My discr: 524295, your discr: 524296, state UP, D/F/P/C/A: 0/0/0/1/0
  Timer Values:
   Local negotiated async tx interval: 2 s
   Remote negotiated async tx interval: 2 s
   Desired echo tx interval: 119 ms, local negotiated echo tx interval: 119 ms
   Echo detection time: 1190 ms(119 ms*10), async detection time: 20 s(2 s*10)
  Local Stats:
   Intervals between async packets:
     Tx: Number of intervals=100, min=1664 ms, max=2001 ms, avg=1838 ms
         Last packet transmitted 314 ms ago
     Rx: Number of intervals=100, min=1662 ms, max=2 s, avg=1828 ms
         Last packet received 1616 ms ago
   Intervals between echo packets:
     Tx: Number of intervals=100, min=120 ms, max=223 ms, avg=125 ms
         Last packet transmitted 112 ms ago
     Rx: Number of intervals=100, min=119 ms, max=223 ms, avg=125 ms
         Last packet received 110 ms ago
   Latency of echo packets (time between tx and rx):
     Number of packets: 100, min=0 us, max=2 ms, avg=850 us
  Session owner information:
    Client           Desired interval         Multiplier
    ---------------- --------------------     --------------
    bgp-0            119 ms                   10
  
  I/f: GigabitEthernet0/3/0/6, Location: 0/3/CPU0, dest: 5.5.5.2, src: 5.5.5.1
   State: UP for 0d:0h:4m:50s, number of times UP: 1
  Received parameters:
   Version: 1, desired tx interval: 2 s, required rx interval: 2 s
   Required echo rx interval: 1 ms, multiplier: 4, diag: None
   My discr: 786436, your discr: 786433, state UP, D/F/P/C/A: 0/0/0/1/0
  Transmitted parameters:
   Version: 1, desired tx interval: 2 s, required rx interval: 2 s
   Required echo rx interval: 1 ms, multiplier: 4, diag: None
   My discr: 786433, your discr: 786436, state UP, D/F/P/C/A: 0/0/0/1/0
  Timer Values:
   Local negotiated async tx interval: 2 s
   Remote negotiated async tx interval: 2 s
   Desired echo tx interval: 318 ms, local negotiated echo tx interval: 318 ms
   Echo detection time: 1272 ms(318 ms*4), async detection time: 8 s(2 s*4)
  Local Stats:
   Intervals between async packets:
     Tx: Number of intervals=100, min=1663 ms, max=2 s, avg=1821 ms
         Last packet transmitted 1740 ms ago
     Rx: Number of intervals=100, min=1663 ms, max=2001 ms, avg=1832 ms
         Last packet received 160 ms ago
   Intervals between echo packets:
     Tx: Number of intervals=100, min=181 ms, max=484 ms, avg=232 ms
         Last packet transmitted 44 ms ago
     Rx: Number of intervals=100, min=179 ms, max=484 ms, avg=232 ms
         Last packet received 41 ms ago
   Latency of echo packets (time between tx and rx):
     Number of packets: 100, min=0 us, max=3 ms, avg=540 us
  Session owner information:
    Client           Desired interval         Multiplier
    ---------------- --------------------     --------------
    bgp-0            318 ms                   4
  
  RP/0/RSP0/CPU0:router# show bgp nei 3.3.3.2
  
  BGP neighbor is 3.3.3.2
   Remote AS 500, local AS 65000, external link
   Remote router ID 16.0.0.1
    BGP state = Established, up for 00:05:01 
    BFD enabled (session up): mininterval: 311 multiplier: 7
    Last read 00:00:56, hold time is 180, keepalive interval is 60 seconds
    Precedence: internet
    Neighbor capabilities:
      Route refresh: advertised and received
      4-byte AS: advertised and received
      Address family IPv4 Unicast: advertised and received
    Received 8 messages, 0 notifications, 0 in queue
    Sent 9 messages, 1 notifications, 0 in queue
    Minimum time between advertisement runs is 30 seconds
   
   For Address Family: IPv4 Unicast
    BGP neighbor version 2
    Update group: 0.2
    AF-dependant capabilities:
      Graceful Restart Capability advertised and received
        Neighbor preserved the forwarding state during latest restart
        Local restart time is 120, RIB purge time is 600 seconds
        Maximum stalepath time is 360 seconds
        Remote Restart time is 120 seconds
    Route refresh request: received 0, sent 0
    Policy for incoming advertisements is pass-all
    Policy for outgoing advertisements is pass-all
    1 accepted prefixes, 1 are bestpaths
    Prefix advertised 1, suppressed 0, withdrawn 0, maximum limit 524288
    Threshold for warning message 75%
    An EoR was not received during read-only mode
  
    Connections established 1; dropped 0      
    Last reset 00:06:58, due to User clear requested (CEASE notification sent - administrative reset)
    Time since last notification sent to neighbor: 00:06:58
    Error Code: administrative reset
    Notification data sent:
      None
  
  RP/0/RSP0/CPU0:router# show bgp nei 5.5.5.2
  
  BGP neighbor is 5.5.5.2
   Remote AS 500, local AS 65000, external link
   Remote router ID 16.0.0.1
    BGP state = Established, up for 00:05:04
    BFD enabled (session up): mininterval: 318 multiplier: 4
    Last read 00:00:58, hold time is 180, keepalive interval is 60 seconds
    Precedence: internet
    Neighbor capabilities:
      Route refresh: advertised and received
      4-byte AS: advertised and received
      Address family IPv4 Unicast: advertised and received
    Received 8 messages, 0 notifications, 0 in queue
    Sent 9 messages, 1 notifications, 0 in queue
    Minimum time between advertisement runs is 30 seconds
  
   For Address Family: IPv4 Unicast
    BGP neighbor version 2
    Update group: 0.2
    AF-dependant capabilities:
      Graceful Restart Capability advertised and received
        Neighbor preserved the forwarding state during latest restart
        Local restart time is 120, RIB purge time is 600 seconds
        Maximum stalepath time is 360 seconds
        Remote Restart time is 120 seconds
    Route refresh request: received 0, sent 0
    Policy for incoming advertisements is pass-all
    Policy for outgoing advertisements is pass-all
    1 accepted prefixes, 0 are bestpaths
    Prefix advertised 1, suppressed 0, withdrawn 0, maximum limit 524288
    Threshold for warning message 75%
    An EoR was not received during read-only mode
  
    Connections established 1; dropped 0
    Last reset 00:07:01, due to User clear requested (CEASE notification sent - administrative reset)
    Time since last notification sent to neighbor: 00:07:01
    Error Code: administrative reset
    Notification data sent:
      None
  
  RP/0/RSP0/CPU0:router# show bgp vrf one nei 3.12.1.2
  
  BGP neighbor is 3.12.1.2, vrf one
   Remote AS 500, local AS 65000, external link
   Remote router ID 16.0.0.1
    BGP state = Established, up for 00:05:06
    BFD enabled (session up): mininterval: 119 multiplier: 10
    Last read 00:00:01, hold time is 180, keepalive interval is 60 seconds
    Precedence: internet
    Neighbor capabilities:
      Route refresh: advertised and received
      4-byte AS: advertised and received
      Address family IPv4 Unicast: advertised and received
    Received 9 messages, 0 notifications, 0 in queue
    Sent 9 messages, 1 notifications, 0 in queue
    Minimum time between advertisement runs is 0 seconds
  
   For Address Family: IPv4 Unicast
    BGP neighbor version 2
    Update group: 0.2
    AF-dependant capabilities:
      Graceful Restart Capability advertised and received
        Neighbor preserved the forwarding state during latest restart
        Local restart time is 120, RIB purge time is 600 seconds
        Maximum stalepath time is 360 seconds
        Remote Restart time is 120 seconds
    Route refresh request: received 0, sent 0
    Policy for incoming advertisements is pass-all
    Policy for outgoing advertisements is pass-all
    1 accepted prefixes, 1 are bestpaths
    Prefix advertised 0, suppressed 0, withdrawn 0, maximum limit 524288
    Threshold for warning message 75%
    An EoR was not received during read-only mode
  
    Connections established 1; dropped 0 
    Last reset 00:07:04, due to User clear requested (CEASE notification sent - administrative reset)
    Time since last notification sent to neighbor: 00:07:04
    Error Code: administrative reset
    Notification data sent:
      None
  
  

bgp as-path-loopcheck

To enable loop checking in the autonomous system path of the prefixes advertised by internal Border Gateway Protocol (iBGP) peers, use the bgp as-path-loopcheck command in an appropriate configuration mode. To restore the system to its default condition, use the no form of this command.

bgp as-path-loopcheck

no bgp as-path-loopcheck

Syntax Description

This command has no keywords or arguments.

Command Default

When you do not specify this command, loop checking is performed only for external peers.

Command Modes

Router configuration

Command History

Release

Modification

Release 3.7.2

This command was introduced.

Usage Guidelines

To use this command, you must be in a user group associated with a task group that includes appropriate task IDs. If the user group assignment is preventing you from using a command, contact your AAA administrator for assistance.

Task ID

Task ID

Operations

bgp

read, write

Examples

The following example shows how to configure an autonomous system path for loop checking iBGP peers:


RP/0/RSP0/CPU0:router(config)# router bgp 6
RP/0/RSP0/CPU0:router(config-bgp)# bgp as-path-loopcheck

bgp attribute-download

To enable Border Gateway Protocol (BGP) attribute download, use the bgp attribute-download command in an appropriate configuration mode. To disable BGP attribute download, use the no form of this command.

bgp attribute-download

no bgp attribute-download

Syntax Description

This command has no keywords or arguments.

Command Default

BGP attribute download is not enabled.

Command Modes

IPv4 unicast address family configuration

Command History

Release

Modification

Release 3.7.2

This command was introduced.

Usage Guidelines

To use this command, you must be in a user group associated with a task group that includes appropriate task IDs. If the user group assignment is preventing you from using a command, contact your AAA administrator for assistance.

When BGP attribute download is enabled using the bgp attribute-download command, BGP reinstalls all routes whose attributes are not currently in the RIB. Likewise, if the user disables BGP attribute download using the no form of the command, BGP reinstalls previously installed routes with a null key, and removes the attributes from the RIB.

Use the bgp attribute-download command to enable the Netflow BGP data export function. When attribute download is enabled, BGP downloads the attribute information for prefixes (community, extended community, and as-path) to the Routing Information Base (RIB) and Forwarding Information Base (FIB). This enables FIB to associate the prefixes with attributes and send the Netflow statistics along with the associated attributes.

Task ID

Task ID

Operations

bgp

read, write

Examples

The following example shows the BGP routes before and after BGP attribute download is enabled and shows how to enable BGP attribute download on BGP router 50:


  RP/0/RSP0/CPU0:router# show route bgp 
  
  B    100.0.1.0/24 [200/0] via 10.0.101.1, 00:00:37
  B    100.0.2.0/24 [200/0] via 10.0.101.1, 00:00:37
  B    100.0.3.0/24 [200/0] via 10.0.101.1, 00:00:37
  B    100.0.4.0/24 [200/0] via 10.0.101.1, 00:00:37
  B    100.0.5.0/24 [200/0] via 10.0.101.1, 00:00:37
  
  RP/0/RSP0/CPU0:router(config)# router bgp 50 
  RP/0/RSP0/CPU0:router(config-bgp)# address-family ipv4 unicast
  RP/0/RSP0/CPU0:router(config-bgp-af)# bgp attribute-download
  !
  !
  !
  RP/0/RSP0/CPU0:router# show route bgp 
  
  B    100.0.1.0/24 [200/0] via 10.0.101.1, 00:00:01
        Attribute ID 0x2
  B    100.0.2.0/24 [200/0] via 10.0.101.1, 00:00:01
        Attribute ID 0x2
  B    100.0.3.0/24 [200/0] via 10.0.101.1, 00:00:01
        Attribute ID 0x2
  B    100.0.4.0/24 [200/0] via 10.0.101.1, 00:00:01
        Attribute ID 0x2
  B    100.0.5.0/24 [200/0] via 10.0.101.1, 00:00:01
        Attribute ID 0x2
  

bgp auto-policy-soft-reset disable

To disable an automatic soft reset of Border Gateway Protocol (BGP) peers when their configured route policy is modified, use the bgp auto-policy-soft-reset disable command in an appropriate configuration mode. To re-enable automatic soft reset of BGP peers, use the no form of this command.

bgp auto-policy-soft-reset disable

no bgp auto-policy-soft-reset disable

Syntax Description

This command has no keywords or arguments.

Command Default

Automatic soft reset of peers is enabled.

Command Modes

Router configuration

VRF configuration

Command History

Release

Modification

Release 3.7.2

This command was introduced.

Usage Guidelines

To use this command, you must be in a user group associated with a task group that includes appropriate task IDs. If the user group assignment is preventing you from using a command, contact your AAA administrator for assistance.


Note


If the inbound policy changes, it is not always possible to perform a soft reset. This is the case if the neighbor does not support route refresh and soft-reconfiguration inbound is not configured for the neighbor. In such instances, a message is logged in the system log indicating that a manual hard reset is needed.


Task ID

Task ID

Operations

bgp

read, write

Examples

The following example shows how to disable an automatic soft reset of BGP peers when their configured route policy is modified:


Router(config)# router bgp 6
Router(config-bgp)# bgp auto-policy-soft-reset disable

bgp bestpath as-path ignore

To ignore the autonomous system path length when calculating preferred paths, use the bgp bestpath as-path ignore command in an appropriate configuration mode. To return the software to the default state in which it considers the autonomous system path length when calculating preferred paths, use the no form of this command.

bgp bestpath as-path ignore

no bgp bestpath as-path ignore

Syntax Description

This command has no keywords or arguments.

Command Default

The autonomous system path length is used (not ignored) when a best path is selected.

Command Modes

Router configuration

VRF configuration

Command History

Release

Modification

Release 3.7.2

This command was introduced.

Usage Guidelines

To use this command, you must be in a user group associated with a task group that includes appropriate task IDs. If the user group assignment is preventing you from using a command, contact your AAA administrator for assistance.

Use the bgp bestpath as-path ignore command to ignore the length of autonomous system paths when the software selects a preferred path. When the best path is selected, if this command is specified, all steps are performed as usual except comparison of the autonomous path length between candidate paths.

Task ID

Task ID

Operations

bgp

read, write

Examples

The following example shows how to configure the software to ignore the autonomous system length when performing best-path selection:


RP/0/RSP0/CPU0:router(config)# router bgp 65000
RP/0/RSP0/CPU0:router(config-bgp)# bgp bestpath as-path ignore

bgp bestpath compare-routerid

To compare identical routes received from external BGP (eBGP) peers during the best-path selection process and select the route with the lowest router ID, use the bgp bestpath compare-routerid command in an appropriate configuration mode. To disable comparing identical routes received from eBGP peers during best-path selection, use the no form of this command.

bgp bestpath compare-routerid

no bgp bestpath compare-routerid

Syntax Description

This command has no keywords or arguments.

Command Default

The software does not select a new best path if it is the same as the current best path (according to the BGP selection algorithm) except for the router ID.

Command Modes

Router configuration

VRF configuration

Command History

Release

Modification

Release 3.7.2

This command was introduced.

Usage Guidelines

To use this command, you must be in a user group associated with a task group that includes appropriate task IDs. If the user group assignment is preventing you from using a command, contact your AAA administrator for assistance.

Use the bgp bestpath compare-routerid command to affect how the software selects the best path, in the case where there are two paths of equal cost according to the BGP selection algorithm. This command is used to force the software to select the path with the lower router ID as the best path. If this command is not used, the software continues to use whichever path is currently the best path, regardless of which has the lower router ID.

Task ID

Task ID

Operations

bgp

read, write

Examples

The following example shows how to configure the BGP speaker in autonomous system 500 to compare the router IDs of similar paths:


RP/0/RSP0/CPU0:router(config)# router bgp 500
RP/0/RSP0/CPU0:router(config-bgp)# bgp bestpath compare-routerid

bgp bestpath cost-community ignore

To configure a router that is running the Border Gateway Protocol (BGP) to not evaluate the cost community attribute during the best-path selection process, use the bgp bestpath cost-community ignore command in an appropriate configuration mode. To restore the system to its default condition, use the no form of this command.

bgp bestpath cost-community ignore

no bgp bestpath cost-community ignore

Syntax Description

This command has no keywords or arguments.

Command Default

The behavior of this command is enabled by default until the cost community attribute is manually configured.

Command Modes

Router configuration

VRF configuration

Command History

Release

Modification

Release 3.7.2

This command was introduced.

Usage Guidelines

To use this command, you must be in a user group associated with a task group that includes appropriate task IDs. If the user group assignment is preventing you from using a command, contact your AAA administrator for assistance.

Use the bgp bestpath cost-community ignore command to disable the evaluation of the cost community attribute to help isolate problems and troubleshoot issues that relate to BGP path selection. This command can also be used to delay the activation of cost community attribute evaluation so that cost community filtering can be deployed in a large network at the same time.

Task ID

Task ID

Operations

bgp

read, write

Examples

The following example shows how to configure a router to not evaluate the cost community attribute during the best-path selection process:


RP/0/RSP0/CPU0:router(config)# router bgp 500
RP/0/RSP0/CPU0:router(config-bgp)# bgp bestpath cost-community ignore

bgp bestpath med always

To allow the comparison of the Multi Exit Discriminator (MED) for paths from neighbors in different autonomous systems, use the bgp bestpath med always command in an appropriate configuration mode. To disable considering the MED attribute in comparing paths, use the no form of this command.

bgp bestpath med always

no bgp bestpath med always

Syntax Description

This command has no keywords or arguments.

Command Default

The software does not compare MEDs for paths from neighbors in different autonomous systems.

Command Modes

Router configuration

VRF configuration

Command History

Release

Modification

Release 3.7.2

This command was introduced.

Usage Guidelines

To use this command, you must be in a user group associated with a task group that includes appropriate task IDs. If the user group assignment is preventing you from using a command, contact your AAA administrator for assistance.

The MED is one of the parameters that is considered by the software when selecting the best path among many alternative paths. The software chooses the path with the lowest MED.

By default, during the best-path selection process, the software makes a MED comparison only among paths from the same autonomous system. This command changes the default behavior of the software by allowing comparison of MEDs among paths regardless of the autonomous system from which the paths are received.

When the bgp bestpath med always command is not enabled and distributed BGP is configured, speakers calculate partial best paths only (executes the best-path steps up to the MED comparison) and send them to BGP Routing Information Base (bRIB). bRIB calculates the final best path (executes all the steps in the best-path calculation). When the bgp bestpath med always command is enabled and distributed BGP is configured, speakers can compare the MED across all ASs, allowing the speaker to calculate a single best path to send it to bRIB. bRIB is the ultimate process that calculates the final best path, but when the bgp bestpath med always command is enabled, the speakers send a single best path instead of potentially sending multiple, partial best paths

Task ID

Task ID

Operations

bgp

read, write

Examples

The following example shows how to configure the Border Gateway Protocol (BGP) speaker in autonomous system 100 to compare MEDs among alternative paths, regardless of the autonomous system from which the paths are received:


RP/0/RSP0/CPU0:router(config)# router bgp 100
RP/0/RSP0/CPU0:router(config-bgp)# bgp bestpath med always

bgp bestpath med confed

To enable Multi Exit Discriminator (MED) comparison among paths learned from confederation peers, use the bgp bestpath med confed command in an appropriate configuration mode. To disable the software from considering the MED attribute in comparing paths, use the no form of this command.

bgp bestpath med confed

no bgp bestpath med confed

Syntax Description

This command has no keywords or arguments.

Command Default

The software does not compare the MED of paths containing only confederation segments, or paths containing confederation segments followed by an AS_SET, with the MED of any other paths.

Command Modes

Router configuration

Command History

Release

Modification

Release 3.7.2

This command was introduced.

Usage Guidelines

To use this command, you must be in a user group associated with a task group that includes appropriate task IDs. If the user group assignment is preventing you from using a command, contact your AAA administrator for assistance.

By default, the MED of the following paths is not compared with the MED of any other path:

  • Paths with an empty autonomous system path
  • Paths beginning with an AS_SET
  • Paths containing only confederation segments
  • Paths containing confederation segments followed by an AS_SET

Use the bgp bestpath med confed command to affect how the following types of paths are treated in the BGP best-path algorithm:

  • Paths containing only confederation segments
  • Paths containing confederation segments followed by an AS_SET

The MED for paths that start with an AS_SEQUENCE or that start with confederation segments followed by an AS_SEQUENCE only is compared with the MED of other paths that share the same first autonomous system number in the autonomous system sequence (the neighbor autonomous system number). This behavior is not affected by the bgp bestpath med confed command.

As an example, suppose that autonomous systems 65000, 65001, 65002, and 65004 are part of a confederation, but autonomous system 1 is not. Suppose that for a particular route, the following paths exist:

  • Path 1: 65000 65004, med = 2, IGP metric = 20
  • Path 2: 65001 65004, med = 3, IGP metric = 10
  • Path 3: 65002 1, med = 1, IGP metric = 30

If the bgp bestpath med confed command is enabled, the software selects path 1 as the best path because it:

  • Has a lower MED than path 2
  • Has a lower IGP metric than path 3

The MED is not compared with path 3 because it has an external autonomous system number (that is, an AS_SEQUENCE) in the path. If the bgp bestpath med confed command is not enabled, then MED is not compared between any of these paths. Consequently, the software selects path 2 as the best path because it has the lowest IGP metric.

Task ID

Task ID

Operations

bgp

read, write

Examples

The following command shows how to enable Border Gateway Protocol (BGP) software to compare MED values for paths learned from confederation peers:


RP/0/RSP0/CPU0:router(config)# router bgp 210
RP/0/RSP0/CPU0:router(config-bgp)# bgp bestpath med confed

bgp bestpath med missing-as-worst

To have the software consider a missing Multi Exit Discriminator (MED) attribute in a path as having a value of infinity, making the path without a MED value the least desirable path, use the bgp bestpath med missing-as-worst command in an appropriate configuration mode. To disable considering the MED attribute in comparing paths, use the no form of this command.

bgp bestpath med missing-as-worst

no bgp bestpath med missing-as-worst

Syntax Description

This command has no keywords or arguments.

Command Default

The software assigns a value of 0 to the missing MED, causing the path with the missing MED attribute to be considered as the best possible MED.

Command Modes

Router configuration

VRF configuration

Command History

Release

Modification

Release 3.7.2

This command was introduced.

Usage Guidelines

To use this command, you must be in a user group associated with a task group that includes appropriate task IDs. If the user group assignment is preventing you from using a command, contact your AAA administrator for assistance.

Task ID

Task ID

Operations

bgp

read, write

Examples

The following example shows how to direct the Border Gateway Protocol (BGP) software to consider a missing MED attribute in a path as having a value of infinity, making this path the least desirable path:


RP/0/RSP0/CPU0:router(config)# router bgp 210
RP/0/RSP0/CPU0:router(config-bgp)# bgp bestpath med missing-as-worst

bgp bestpath origin-as allow invalid

To permit all paths marked with an 'invalid' origin-as by RPKI to be considered for BGP best path computation, use the bgp bestpath origin-as allow invalid command in the router configuration mode. This configuration can also be made in global address family, neighbor, and neighbor address family submodes. To return the device to default operation, use the no form of this command.

For router and global address family configuration mode:

bgp bestpath origin-as allow invalid

no bgp bestpath origin-as allow invalid

For neighbour and neighbor address family configuration mode:

bestpath origin-as allow invalid

no bestpath origin-as allow invalid

Syntax Description

This command has no keywords or arguments.

Command Default

By default, prefixes marked with an 'invalid' origin-as are not considered for BGP best path computation when the router is performing origin-as validation.

Command Modes

Router configuration

Address family

Neighbor

Neighbor Address family

Command History

Release Modification
Release 4.2.1

This command was introduced

Usage Guidelines

To use this command, you must be in a user group associated with a task group that includes appropriate task IDs. If the user group assignment is preventing you from using a command, contact your AAA administrator for assistance.

Configuring the bgp bestpath origin-as allow invalid command allows paths marked with an 'invalid' origin-as to be considered for best path computation. This can be limited to an address family by configuring it at the address-family submode.

This configuration takes effect only when the bgp bestpath origin-as use validity configuration is enabled.

Task ID

Task ID Operation

bgp

read, write

Examples

The following example shows how to permit all invalid paths to be considered for BGP best-path selection in the global mode:

RP/0/RSP0/CPU0:router#configure
RP/0/RSP0/CPU0:router(config)#router bgp 50000
RP/0/RSP0/CPU0:router(config-bgp)#bgp bestpath origin-as allow invalid

The following example shows how to permit all invalid paths to be considered for BGP best-path selection in the address family submode:

RP/0/RSP0/CPU0:router#configure
RP/0/RSP0/CPU0:router(config)#router bgp 50000
RP/0/RSP0/CPU0:router(config-bgp)#address-family ipv4 unicast
RP/0/RSP0/CPU0:router(config-bgp-af)#bgp bestpath origin-as allow invalid 

The following example shows how to permit all invalid paths to be considered for best-path selection in the neighbor submode:

RP/0/RSP0/CPU0:router#configure
RP/0/RSP0/CPU0:router(config)#router bgp 50000
RP/0/RSP0/CPU0:router(config-bgp)#neighbor 1.1.1.1
RP/0/RSP0/CPU0:router(config-bgp-nbr)#bestpath origin-as allow invalid 

The following example shows how to permit all invalid paths to be considered for best-path selection in the neighbour address-family submode:

RP/0/RSP0/CPU0:router#configure
RP/0/RSP0/CPU0:router(config)#router bgp 50000
RP/0/RSP0/CPU0:router(config-bgp)#neighbor 1.1.1.1
RP/0/RSP0/CPU0:router(config-bgp-nbr)#address-family ipv4 unicast
RP/0/RSP0/CPU0:router(config-bgp-nbr-af)#bestpath origin-as allow invalid 

bgp bestpath origin-as use validity

To enable the BGP Origin AS Validation feature (RPKI) and allow the validity states of BGP paths to be taken into consideration in the bestpath process, use the bgp bestpath origin-as use validity command. This can be configured in router configuration mode and address family submode. To return the device to default operation, use the no form of this command.

bgp bestpath origin-as use validity

no bgp bestpath origin-as use validity

Syntax Description

This command has no keywords or arguments.

Command Default

By default, the best path computation does not take RPKI states into account.

Command Modes

Router configuration

Address family configuration

Command History

Release Modification
Release 4.2.1

This command was introduced

Usage Guidelines

To use this command, you must be in a user group associated with a task group that includes appropriate task IDs. If the user group assignment is preventing you from using a command, contact your AAA administrator for assistance.

There are three RPKI states - valid, invalid, and not found. When the bgp bestpath origin-as use validity command is configured, only paths marked with 'valid' or 'not found' are considered as best path candidates. When the bgp bestpath origin-as allow invalid command is configured, paths marked as 'invalid' are also considered but preference is given to routes marked 'valid' over those marked 'invalid'.

The bgp bestpath origin-as use validity command takes precedence over the configured weight and local preferences attributes.

Task ID

Task ID Operation

bgp

read, write

Examples

The following example shows how to enable the validity states of BGP paths to affect the path's preference when performing best-path selection:

RP/0/RSP0/CPU0:router#configure
RP/0/RSP0/CPU0:router(config)#router bgp 50000
RP/0/RSP0/CPU0:router(config-bgp)#bgp bestpath origin-as use validity

bgp bestpath aigp ignore

To configure a device that is running the Border Gateway Protocol (BGP) to not evaluate the accumulated interior gateway protocol (AIGP) metric during the best path selection process between two paths when one path does not have the AIGP metric, use the bgp bestpath aigp ignore command in router configuration mode. To return the device to default operation, use the no form of this command.

bgp bestpath aigp ignore

no bgp bestpath aigp ignore

Syntax Description

This command has no keywords or arguments.

Command Default

AIGP is enabled by default.

If this command is not configured, then the accumulated interior gateway protocol (AIGP) metric is evaluated (not ignored) during the best path selection.

Command Modes

Router configuration

VRF configuration

Command History

Release Modification
Release 4.1.1

This command was introduced.

Usage Guidelines

To use this command, you must be in a user group associated with a task group that includes appropriate task IDs. If the user group assignment is preventing you from using a command, contact your AAA administrator for assistance.

By default, BGP always prefers a path with the AIGP metric. When there are two paths, one with the AIGP metric and the other without, then executing the bgp bestpath aigp ignore command results in BGP performing best path computation as if neither paths has the AIGP metric.

Task ID

Task ID

Operations

bgp

read, write

Examples

The following example shows how to configure the software to ignore the accumulated interior gateway protocol (AIGP) metric when performing best-path selection:

RP/0/RSP0/CPU0:router#configure
RP/0/RSP0/CPU0:router(config)#router bgp 50000
RP/0/RSP0/CPU0:router(config-bgp)#bgp bestpath aigp ignore

bgp bestpath as-path multipath-relax

To configure a Border Gateway Protocol (BGP) routing process to consider the different autonomous system (AS) paths and load balance multiple paths during best path route selection, use the bgp bestpath as-path multipath-relax command. To return the BGP routing process to the default operation, use the no form of this command.

bgp bestpath as-path multipath-relax

no bgp bestpath as-path multipath-relax

Syntax Description

This command has no keywords or arguments.

Command Default

None

Command Modes

Router BGP configuration

VRF configuration

Command History

Release Modification
Release 3.7.2

This command was introduced.

Usage Guidelines

To use this command, you must be in a user group associated with a task group that includes appropriate task IDs. If the user group assignment is preventing you from using a command, contact your AAA administrator for assistance.

When BGP multi-pathing is enabled, BGP load-balances user traffic within a single autonomous system (AS). The criteria are that all attributes must match (weight, AS path, etc). However when a device is multi-homed to multiple autonomous systems, BGP cannot load balance traffic between them by default. In order to enable load-balancing of traffic among the multi-homed autonomous systems, the bgp bestpath as-path multipath-relax command needs to be enabled. The criteria required for this is that the AS-path length should be equal.

Before you use this command, ensure that BGP is enabled

Task ID

Task ID Operation

bgp

read, write

Examples

This example shows how to configure multipath load sharing on paths from different autonomous systems in router mode:

RP/0/RSP0/CPU0:router#configure
RP/0/RSP0/CPU0:router(config)#router bgp 120
RP/0/RSP0/CPU0:router(config-bgp)#bgp bestpath as-path multipath-relax

bgp client-to-client reflection disable

To disable reflection of routes between route-reflection clients using a Border Gateway Protocol (BGP) route reflector, use the bgp client-to-client reflection disable command in address family configuration mode. To re-enable client-to-client reflection, use the no form of this command.

bgp client-to-client reflection [cluster-id cluster-id] disable

no bgp client-to-client reflection [cluster-id cluster-id] disable

Syntax Description

This command has no keywords or arguments.

Command Default

Client-to-client reflection is enabled.

Command Modes

Address family configuration

Command History

Release

Modification

Release 3.7.2

This command was introduced.

Usage Guidelines

To use this command, you must be in a user group associated with a task group that includes appropriate task IDs. If the user group assignment is preventing you from using a command, contact your AAA administrator for assistance.

By default, the clients of a route reflector are not required to be fully meshed and the routes from a client are reflected to other clients. However, if the clients are fully meshed, route reflection is not required.

Examples

In this example, the three neighbors are fully meshed, so client-to-client reflection is disabled:


RP/0/RSP0/CPU0:router(config)# router bgp 65534
RP/0/RSP0/CPU0:router(config-bgp)# address-family ipv4 unicast 
RP/0/RSP0/CPU0:router(config-bgp-af)# bgp client-to-client reflection disable 
RP/0/RSP0/CPU0:router(config-bgp-af)# exit
RP/0/RSP0/CPU0:router(config-bgp)# neighbor-group rrclients
RP/0/RSP0/CPU0:router(config-bgp-nbrgrp)# remote-as 65534
RP/0/RSP0/CPU0:router(config-bgp-nbrgrp)# address-family ipv4 unicast 
RP/0/RSP0/CPU0:router(config-bgp-nbrgrp-af)# route-reflector-client 
RP/0/RSP0/CPU0:router(config-bgp-nbrgrp-af)# exit
RP/0/RSP0/CPU0:router(config-bgp-nbrgrp)# exit

RP/0/RSP0/CPU0:router(config-bgp)# neighbor 192.168.253.21 use neighbor-group rrclients
RP/0/RSP0/CPU0:router(config-bgp)# neighbor 192.168.253.22 use neighbor-group rrclients

bgp cluster-id

To configure the cluster ID if the Border Gateway Protocol (BGP) cluster has more than one route reflector, use the bgp cluster-id command in an appropriate configuration mode. To remove the cluster ID, use the no form of this command.

bgp cluster-id cluster-id

no bgp cluster-id [cluster-id]

Syntax Description

cluster-id

Cluster ID of this router acting as a route reflector; maximum of 4 bytes. Cluster ID can be entered either as an IP address or value. Range is 1 to 4294967295.

Command Default

A cluster ID is not configured.

Command Modes

Router configuration

Command History

Release

Modification

Release 3.7.2

This command was introduced.

Usage Guidelines

To use this command, you must be in a user group associated with a task group that includes appropriate task IDs. If the user group assignment is preventing you from using a command, contact your AAA administrator for assistance.

Together, a route reflector and its clients form a cluster. A cluster of clients usually has a single route reflector. In such instances, the cluster is identified by the software as the router ID of the route reflector. To increase redundancy and avoid a single point of failure in the network, a cluster might have more than one route reflector. If it does, all route reflectors in the cluster must be configured with the same 4-byte cluster ID so that a route reflector can recognize updates from route reflectors in the same cluster.

Task ID

Task ID

Operations

bgp

read, write

Examples

The following example shows how to configure the local router as one of the route reflectors serving the cluster. Neighbor 192.168.70.24 is assigned to the default cluster with cluster-id 1.


RP/0/RSP0/CPU0:router(config)# router bgp 65534
RP/0/RSP0/CPU0:router(config-bgp)# bgp cluster-id 1
RP/0/RSP0/CPU0:router(config-bgp)# neighbor 192.168.70.24
RP/0/RSP0/CPU0:router(config-bgp-nbr)# remote-as 65534
RP/0/RSP0/CPU0:router(config-bgp-nbr)# address-family ipv4 unicast  
RP/0/RSP0/CPU0:router(config-bgp-nbr-af)# route-reflector-client 

bgp confederation identifier

To specify a Border Gateway Protocol (BGP) confederation identifier, use the bgp confederation identifier command in an appropriate configuration mode. To remove the confederation identifier, use the no form of this command.

bgp confederation identifier as-number

no bgp confederation identifier [as-number]

Syntax Description

as-number

Autonomous system (AS) number that internally includes multiple autonomous systems.

  • Range for 2-byte Autonomous system numbers (ASNs) is 1 to 65535.
  • Range for 4-byte Autonomous system numbers (ASNs) in asplain format is 1 to 4294967295.
  • Range for 4-byte Autonomous system numbers (ASNs) is asdot format is 1.0 to 65535.65535.

Command Default

No confederation identifier is configured.

Command Modes

Router configuration

Command History

Release

Modification

Release 3.7.2

This command was introduced.

Release 3.9.0

Asplain format for 4-byte Autonomous system numbers notation was supported.

Usage Guidelines

To use this command, you must be in a user group associated with a task group that includes appropriate task IDs. If the user group assignment is preventing you from using a command, contact your AAA administrator for assistance.

One way to reduce the internal BGP (iBGP) mesh is to divide an autonomous system into multiple autonomous systems and group them into a single confederation. Each autonomous system is fully meshed within itself, and has a few connections to another autonomous system in the same confederation. Although the peers in different autonomous systems have external BGP (eBGP) sessions, they exchange routing information as if they are iBGP peers. Specifically, the confederation maintains the next hop and local preference information, and that allows you to retain a single Interior Gateway Protocol (IGP) for all autonomous systems. To the outside world, the confederation looks like a single autonomous system.

Use the bgp confederation identifier command to specify the autonomous system number for the confederation. This autonomous system number is used when BGP sessions are established with external peers in autonomous systems that are not part of the confederation.

Task ID

Task ID

Operations

bgp

read, write

Examples

The following example shows how to divide the autonomous system into autonomous systems 4001, 4002, 4003, 4004, 4005, 4006, and 4007 with the confederation identifier 5. Neighbor 10.2.3.4 is a router inside the confederation. Neighbor 172.20.16.6 is outside the routing domain confederation. To the outside world, there appears to be a single autonomous system with the number 5.


RP/0/RSP0/CPU0:router(config)# router bgp 4001
RP/0/RSP0/CPU0:router(config-bgp)# bgp confederation identifier 5
RP/0/RSP0/CPU0:router(config-bgp)# bgp confederation peers 4002
RP/0/RSP0/CPU0:router(config-bgp)# bgp confederation peers 4003
RP/0/RSP0/CPU0:router(config-bgp)# bgp confederation peers 4004
RP/0/RSP0/CPU0:router(config-bgp)# bgp confederation peers 4005
RP/0/RSP0/CPU0:router(config-bgp)# bgp confederation peers 4006
RP/0/RSP0/CPU0:router(config-bgp)# bgp confederation peers 4007
RP/0/RSP0/CPU0:router(config-bgp)# neighbor 10.2.3.4
RP/0/RSP0/CPU0:router(config-bgp-nbr)# remote-as 4002 
RP/0/RSP0/CPU0:router(config-bgp-nbr)# exit
RP/0/RSP0/CPU0:router(config-bgp)# exit
RP/0/RSP0/CPU0:router(config-bgp-nbr)# neighbor 172.20.16.6 
RP/0/RSP0/CPU0:router(config-bgp-nbr)# remote-as 4009

bgp confederation peers

To configure the autonomous systems that belong to the confederation, use the bgp confederation peers command in an appropriate configuration mode. To remove the autonomous system from the confederation, use the no form of this command.

bgp confederation peers [as-number]

no bgp confederation peers [as-number]

Syntax Description

as-number

Autonomous system (AS) numbers for Border Gateway Protocol (BGP) peers that belong to the confederation.

  • Range for 2-byte Autonomous system numbers (ASNs) is 1 to 65535.
  • Range for 4-byte Autonomous system numbers (ASNs) in asplain format is 1 to 4294967295.
  • Range for 4-byte Autonomous system numbers (ASNs) is asdot format is 1.0 to 65535.65535.

Command Default

No BGP peers are identified as belonging to the confederation.

Command Modes

Router configuration

Command History

Release

Modification

Release 3.7.2

This command was introduced.

Release 3.9.0

Asplain format for 4-byte Autonomous system numbers notation was supported.

Usage Guidelines

To use this command, you must be in a user group associated with a task group that includes appropriate task IDs. If the user group assignment is preventing you from using a command, contact your AAA administrator for assistance.

The autonomous systems specified in this command are visible internally to a confederation. Each autonomous system is fully meshed within itself. The bgp confederation identifier command specifies the confederation to which the autonomous systems belong.

To specify multiple autonomous systems, enter BGP confederation peer configuration mode then enter one autonomous-system-number for each command line.

Task ID

Task ID

Operations

bgp

read, write

Examples

The following example shows that autonomous systems 1090 and 1093 belong to a single confederation:

RP/0/RSP0/CPU0:router(config)# router bgp 1090
RP/0/RSP0/CPU0:router(config-bgp)# bgp confederation peers 1093

The following example shows that autonomous systems 1095, 1096, 1097, and 1098 belong to a single confederation:


RP/0/RSP0/CPU0:router(config)# router bgp 1095
RP/0/RSP0/CPU0:router(config-bgp)# bgp confederation peers
RP/0/RSP0/CPU0:router(config-bgp-confed-peers)# 1096
RP/0/RSP0/CPU0:router(config-bgp-confed-peers)# 1097
RP/0/RSP0/CPU0:router(config-bgp-confed-peers)# 1098

bgp dampening

To enable Border Gateway Protocol (BGP) route dampening or change various BGP route dampening factors, use the bgp dampening command in an appropriate configuration mode. To disable route dampening and reset default values, use the no form of this command.

bgp dampening [half-life [reuse suppress max-suppress-time] | route-policy route-policy-name]

no bgp dampening [half-life [reuse suppress max-suppress-time] | route-policy route-policy-name]

Syntax Description

half-life

(Optional) Time (in minutes) after which a penalty is decreased. Once the route has been assigned a penalty, the penalty is decreased by half after the half-life period (which is 15 minutes by default). Penalty reduction happens every 5 seconds. Range of the half-life period is from 1 to 45 minutes.

reuse

(Optional) Value for route reuse if the flapping route penalty decreases and falls below the reuse value. When this happens, the route is unsuppressed. The process of unsuppressing routes occurs at 10-second increments. Range is 1 to 20000.

suppress

(Optional) Maximum penalty value. Suppress a route when its penalty exceeds the value specified. When this happens, the route is suppressed. Range is 1 to 20000.

max-suppress-time

(Optional) Maximum time (in minutes) a route can be suppressed. Range is 1 to 255. If the half-life value is allowed to default, the maximum suppress time defaults to 60 minutes.

route-policy route-policy-name

(Optional) Specifies the route policy to use to set dampening parameters.

Command Default

Route dampening is disabled.

half-life : 15 minutes

reuse : 750

suppress : 2000

max-suppress-time : four times half-life value

Command Modes

IPv4 address family configuration

IPv6 address family configuration

VPNv4 address family configuration

VRF IPv4 address family configuration

VPNv6 address family configuration

VRF IPv6 address family configuration

Command History

Release

Modification

Release 3.7.2

This command was introduced.

Usage Guidelines

To use this command, you must be in a user group associated with a task group that includes appropriate task IDs. If the user group assignment is preventing you from using a command, contact your AAA administrator for assistance.

Use the bgp dampening command without arguments to enable BGP route dampening with the default parameters. The parameters can be changed by setting them on the command line or specifying them with a routing policy.

Task ID

Task ID

Operations

bgp

read, write

Examples

The following example shows how to set the half-life value to 30 minutes, the reuse value to 1500, the suppress value to 10000, and the max-suppress-time to 120 minutes:


Router(config)# router bgp 50
Router(config-bgp)# address-family ipv4 unicast
Router(config-bgp-af)# bgp dampening 30 1500 10000 120

bgp default local-preference

To change the default local preference value, use the bgp default local-preference command in an appropriate configuration mode. To reset the local preference value to the default of 100, use the no form of this command.

bgp default local-preference value

no bgp default local-preference [value]

Syntax Description

value

Local preference value. Range is 0 to 4294967295. Higher values are preferable.

Command Default

Enabled with a value of 100.

Command Modes

Router configuration

VRF configuration

Command History

Release

Modification

Release 3.9.0

This command was introduced.

Usage Guidelines

To use this command, you must be in a user group associated with a task group that includes appropriate task IDs. If the user group assignment is preventing you from using a command, contact your AAA administrator for assistance.

Generally, the default value of 100 allows you to easily define a particular path as less preferable than paths with no local preference attribute. The preference is sent to all networking devices in the local autonomous system.

Task ID

Task ID

Operations

bgp

read, write

Examples

The following example shows how to raise the default local preference value from the default of 100 to 200:

Router(config)# router bgp 200
Router(config-bgp)# bgp default local-preference 200

bgp enforce-first-as disable

To disable the software from enforcing the first autonomous system path (known as the AS path) of a route received from an external Border Gateway Protocol (eBGP) peer to be the same as the configured remote autonomous system, use the bgp enforce-first-as disable command in an appropriate configuration mode. To re-enable enforcing the first AS path of a received route from an eBGP peer to be the same as the remote autonomous system, use the no form of this command.

bgp enforce-first-as disable

no bgp enforce-first-as disable

Syntax Description

This command has no keywords or arguments.

Command Default

By default, the software requires the first autonomous system (in the AS path) of a route received from an eBGP peer to be the same as the remote autonomous system configured.

Command Modes

Router configuration

VRF configuration

Command History

Release

Modification

Release 3.7.2

This command was introduced.

Usage Guidelines

To use this command, you must be in a user group associated with a task group that includes appropriate task IDs. If the user group assignment is preventing you from using a command, contact your AAA administrator for assistance.

By default, the software ignores any update received from an eBGP neighbor that does not have the autonomous system configured for that neighbor at the beginning of the AS path. When configured, the command applies to all eBGP peers of the router.

Task ID

Task ID

Operations

bgp

read, write

Examples

The following example shows a configuration in which incoming updates from eBGP neighbors are not checked to ensure the first AS number in the AS path is the same as the configured AS number for the neighbor:


RP/0/RSP0/CPU0:router(config)# router bgp 100
RP/0/RSP0/CPU0:router(config-bgp)# bgp enforce-first-as disable

bgp fast-external-fallover disable

To disable immediately resetting the Border Gateway Protocol (BGP) sessions of any directly adjacent external peers if the link used to reach them goes down, use the bgp fast-external-fallover disable command in an appropriate configuration mode. To disable this function and perform an immediate reset of BGP sessions when a link between peers is lost, use the no form of this command.

bgp fast-external-fallover disable

no bgp fast-external-fallover disable

Syntax Description

disable

Disables BGP fast external failover.

Command Default

BGP sessions of any directly adjacent external peers are immediately reset if the link used to reach them goes down.

Command Modes

Router configuration

VRF configuration

Command History

Release

Modification

Release 3.7.2

This command was introduced.

Usage Guidelines

To use this command, you must be in a user group associated with a task group that includes appropriate task IDs. If the user group assignment is preventing you from using a command, contact your AAA administrator for assistance.

By default, BGP sessions of any directly adjacent external peers are immediately reset, which allows the network to recover faster when links go down between BGP peers.

Task ID

Task ID

Operations

bgp

read, write

Examples

The following example shows how to disable the automatic resetting of BGP sessions:


RP/0/RSP0/CPU0:router(config)# router bgp 109
RP/0/RSP0/CPU0:router(config-bgp)# bgp fast-external-fallover disable

bgp graceful-restart

To enable graceful restart support, use the bgp graceful-restart command in an appropriate configuration mode. To disable this function, use the no form of this command.

bgp graceful-restart

no bgp graceful-restart

Syntax Description

This command has no keywords or arguments.

Command Default

Graceful restart support is not enabled.

Command Modes

Router configuration

Command History

Release

Modification

Release 3.7.2

This command was introduced.

Usage Guidelines

To use this command, you must be in a user group associated with a task group that includes appropriate task IDs. If the user group assignment is preventing you from using a command, contact your AAA administrator for assistance.

Use the bgp graceful-restart command to enable graceful restart functionality on the router, and also to advertise graceful restart to neighboring routers.


Note


The bgp graceful-restart command with no options must be used to enable graceful restart before using the bgp graceful-restart purge-time , bgp graceful-restart restart-time , bgp graceful-restart stalepath-time , or bgp graceful-restart graceful-reset commands.


When graceful restart is enabled, the BGP graceful restart capability is negotiated with neighbors in the BGP OPEN message when the session is established. If the neighbor also advertises support for graceful restart, then graceful restart is activated for that neighbor session. If the neighbor does not advertise support for graceful restart, then graceful restart is not activated for that neighbor session even though it is enabled locally.

If you enter the bgp graceful-restart command after some BGP sessions are established, you must restart those sessions before graceful restart takes effect. Use the clear bgp command to restart sessions.

Unconfiguring a neighbor (no neighbor command) or BGP (no router bgp command) may trigger a graceful restart (GR) on the neighboring router. If you do not want the BGP to perform a graceful restart, disable the BGP neighbor by configuring an administrative shutdown under the neighbor settings before unconfiguring the neighbor or BGP.

Task ID

Task ID

Operations

bgp

read, write

Examples

The following example shows how to enable graceful restart:


RP/0/RSP0/CPU0:router(config)#router bgp 3 
RP/0/RSP0/CPU0:router(config-bgp)#bgp graceful-restart 

bgp graceful-restart graceful-reset

To invoke a graceful restart when configuration changes force a peer reset, use the bgp graceful-restart graceful-reset command in an appropriate configuration mode. To disable this function, use the no form of this command.

bgp graceful-restart graceful-reset

no bgp graceful-restart graceful-reset

Syntax Description

This command has no keywords or arguments.

Command Default

Graceful restart is not invoked when a configuration change forces a peer reset.

Command Modes

Router configuration

Command History

Release

Modification

Release 3.7.2

This command was introduced.

Usage Guidelines

To use this command, you must be in a user group associated with a task group that includes appropriate task IDs. If the user group assignment is preventing you from using a command, contact your AAA administrator for assistance.

BGP graceful restart must be enabled using the bgp graceful-restart command before enabling graceful reset using the bgp graceful-restart graceful-reset command.

Task ID

Task ID

Operations

bgp

read, write

Examples

The following example shows how to enable graceful reset:


RP/0/RSP0/CPU0:router(config)#router bgp 3 
RP/0/RSP0/CPU0:router(config-bgp)# bgp graceful-restart graceful-reset

bgp graceful-restart purge-time

To specify the maximum time before stale routes are purged from the routing information base (RIB) when the local BGP process restarts, use the bgp graceful-restart purge-time command in an appropriate configuration mode. To set the purge timer time to its default value, use the no form of this command.

bgp graceful-restart purge-time seconds

no bgp graceful-restart purge-time seconds

Syntax Description

seconds

Maximum time before stale routes are purged. Time in seconds. Range is 0 to 6000.

Command Default

seconds : 600

Command Modes

Router configuration

Command History

Release

Modification

Release 3.7.2

This command was introduced.

Usage Guidelines

To use this command, you must be in a user group associated with a task group that includes appropriate task IDs. If the user group assignment is preventing you from using a command, contact your AAA administrator for assistance.

BGP graceful restart must be enabled using the bgp graceful-restart command before setting the purge time using the bgp graceful-restart purge-time command.

Task ID

Task ID

Operations

bgp

read, write

Examples

The following example shows how to change the BGP purge time to 800 seconds:


RP/0/RSP0/CPU0:router(config)# router bgp 3
RP/0/RSP0/CPU0:router(config-bgp)# bgp graceful-restart purge-time 800

bgp graceful-restart restart-time

To specify a user-predicted local BGP process maximum restart time, which is advertised to neighbors during session establishment, use the bgp graceful-restart restart-time command in an appropriate configuration mode. To set this restart time to its default value, use the no form of this command.

bgp graceful-restart restart-time seconds

no bgp graceful-restart restart-time seconds

Syntax Description

seconds

Maximum time advertised to neighbors. Time in seconds. Range is 1 to 4095.

Command Default

seconds : 120

Command Modes

Router configuration

Command History

Release

Modification

Release 3.7.2

This command was introduced.

Usage Guidelines

To use this command, you must be in a user group associated with a task group that includes appropriate task IDs. If the user group assignment is preventing you from using a command, contact your AAA administrator for assistance.

BGP graceful restart must be enabled using the bgp graceful-restart command before setting the restart timer using the bgp graceful-restart restart-time command.

Task ID

Task ID

Operations

bgp

read, write

Examples

The following example shows how to change the BGP graceful restart time to 400 seconds:


RP/0/RSP0/CPU0:router(config)#router bgp 3 
RP/0/RSP0/CPU0:router(config-bgp)# bgp graceful-restart restart-time 400

bgp graceful-restart stalepath-time

To specify the maximum time to wait for an End-of-RIB message after a neighbor restarts, use the bgp graceful-restart stalepath-time command in an appropriate configuration mode. To set the stalepath timer time to its default value, use the no form of this command.

bgp graceful-restart stalepath-time seconds

no bgp graceful-restart stalepath-time seconds

Syntax Description

seconds

Maximum wait time. Time in seconds. Range is 1 to 4095.

Command Default

seconds : 360

Command Modes

Router configuration

Command History

Release

Modifications

Release 3.7.2

This command was introduced.

Usage Guidelines

To use this command, you must be in a user group associated with a task group that includes appropriate task IDs. If the user group assignment is preventing you from using a command, contact your AAA administrator for assistance.

BGP graceful restart must be enabled using the bgp graceful-restart command before setting the stalepath time using the bgp graceful-restart stalepath-time command.

If the stalepath time is exceeded before an End-of-RIB message is received from a neighbor, paths learned from the neighbor are purged from the BGP routing table.

Task ID

Task ID

Operations

bgp

read, write

Examples

The following example shows how to change the stalepath time to 750 seconds:


RP/0/RSP0/CPU0:router(config)# router bgp 3
RP/0/RSP0/CPU0:router(config-bgp)# bgp graceful-restart stalepath-time 750

bgp import-delay

To enable delay for Border Gateway Protocol (BGP) batch import processing, use the bgp import-delay command in an appropriate configuration mode. To disable delay in batch import processing, use the no form of this command.

bgp import-delay seconds milliseconds

no bgp import-delay

Syntax Description

seconds

Specifies batch import processing delay in seconds. Range is 0 to 10 seconds.

milliseconds

Specifies batch import processing delay in milliseconds. Range is 0 to 999 seconds.

Command Default

No delay is configured.

Command Modes

Address-family VPNv4 Unicast

Address-family VPNv6 Unicast

Command History

Release Modification
Release 3.9.1

This command was introduced.

Usage Guidelines

To use this command, you must be in a user group associated with a task group that includes appropriate task IDs. If the user group assignment is preventing you from using a command, contact your AAA administrator for assistance.

Task ID

Task ID Operation

bgp

read, write

Examples

This example shows how to set delay in batch import processing as two seconds and zero milliseconds:


RP/0/RSP0/CPU0:router#configure
RP/0/RSP0/CPU0:router(config)#router bgp 100
RP/0/RSP0/CPU0:router(config-bgp)#address-family vpnv4 unicast 
RP/0/RSP0/CPU0:router(config-bgp-af)#bgp import-delay 2 0

bgp label-delay

To enable delay for Border Gateway Protocol (BGP) batch label processing, use the bgp label-delay command in an appropriate configuration mode. To disable delay in batch import processing, use the no form of this command.

bgp label-delay seconds milliseconds

no bgp label-delay

Syntax Description

seconds

Specifies batch label processing delay in seconds. Range is 0 to 10 seconds.

milliseconds

Specifies batch label processing delay in milliseconds. Range is 0 to 999 seconds.

Command Default

No delay is configured.

Command Modes

Address-family IPv4 Unicast

Address-family IPv6 Unicast

Address-family IPv4 Multicast

Address-family IPv6 Multicast

Address-family VPNv4 Unicast

Address-family VPNv6 Unicast

Command History

Release Modification
Release 3.9.1

This command was introduced.

Usage Guidelines

To use this command, you must be in a user group associated with a task group that includes appropriate task IDs. If the user group assignment is preventing you from using a command, contact your AAA administrator for assistance.

Task ID

Task ID Operation

bgp

read, write

Examples

This example shows how to set delay in batch import processing as two seconds and zero milliseconds:


RP/0/RSP0/CPU0:router#configure
RP/0/RSP0/CPU0:router(config)#router bgp 100
RP/0/RSP0/CPU0:router(config-bgp)#address-family ipv4 unicast
RP/0/RSP0/CPU0:router(config-bgp-af)#bgp label-delay 2 0

bgp log neighbor changes disable

To disable logging of Border Gateway Protocol (BGP) neighbor resets, use the bgp log neighbor changes disable command in an appropriate configuration mode. To re-enable logging of BGP neighbor resets, use the no form of this command.

bgp log neighbor changes disable

no bgp log neighbor changes disable

Syntax Description

This command has no keywords or arguments.

Command Default

BGP neighbor changes are logged.

Command Modes

Router configuration

VRF configuration

Command History

Release

Modification

Release 3.7.2

This command was introduced.

Usage Guidelines

To use this command, you must be in a user group associated with a task group that includes appropriate task IDs. If the user group assignment is preventing you from using a command, contact your AAA administrator for assistance.

Logging of BGP neighbor status changes (up or down) and resets is used for troubleshooting network connectivity problems and measuring network stability. Unexpected neighbor resets might indicate high error rates or high packet loss in the network, and should be investigated.

Status change message logging does not substantially affect performance, unlike, for example, enabling per-BGP update debugging. If the UNIX syslog facility is enabled, messages are sent by the software to the UNIX host running the syslog daemon so that the messages can be stored and archived on disk. If the UNIX syslog facility is not enabled, the status change messages are kept in the internal buffer of the router, and are not stored to disk.

The neighbor status change messages are not tracked if the bgp log neighbor changes disable command is disabled, except for the last reset reason, which is always available as output of the show bgp neighbors command.

Up and down messages for BGP neighbors are logged by the software by default. Use the bgp log neighbor changes disable command to stop logging BGP neighbor changes.

Task ID

Task ID

Operations

bgp

read, write

Examples

The following example shows how to prevent the logging of neighbor changes for BGP:


RP/0/RSP0/CPU0:router(config)# router bgp 65530
RP/0/RSP0/CPU0:router(config-bgp)# bgp log neighbor changes disable

bgp lpts-secure-binding

To enable Local Packet Transport Services (LPTS) secure binding, use the bgp lpts-secure-binding command in BGP configuration mode. To disable the LPTS secure binding, use the no form of this command.

bgp lpts-secure-binding

Syntax Description

This command has no arguments or keywords.

Command Default

LPTS secure binding is not enabled.

Command Modes

Router BGP Configuration

Command History

Release Modification
Release 7.10.1

This command was introduced.

Usage Guidelines

None.

Examples

This example shows how to configure LPTS secure binding:

Router# router bgp 100
 Router(config-bgp)# bgp lpts-secure-binding

bgp maximum neighbor

To control the maximum number of neighbors that can be configured on the router, use the bgp maximum neighbor command in an appropriate configuration mode. To set the neighbor limit to the default value, use the no form of this command.

bgp maximum neighbor limit

no maximum neighbor [limit]

Syntax Description

limit

Maximum number of neighbors. Range is 1 to 15000.

Command Default

Default limit is 10000.

Command Modes

Router configuration

Command History

Release

Modification

Release 3.7.2

This command was introduced.

Usage Guidelines

To use this command, you must be in a user group associated with a task group that includes appropriate task IDs. If the user group assignment is preventing you from using a command, contact your AAA administrator for assistance.

Any attempt to configure the neighbor limit below 1 or above 15000 fails. Similarly, attempting to configure the limit below the number of neighbors currently configured fails. For example, if there are 3250 neighbors configured, you cannot set the limit below 3250.

Task ID

Task ID

Operations

bgp

write

Examples

The following example shows how to change the default maximum neighbor limit and set it to 1200:


RP/0/RSP0/CPU0:router(config)#router bgp 65530 
RP/0/RSP0/CPU0:router(config-bgp)# bgp maximum neighbor 1200

bgp multipath as-path

To ignore as-path onwards while computing multipath, use the bgp multipath as-path command in router configuration mode.

bgp multipath as-path ignore onwards

Syntax Description

ignore

Ignores as-path related check for multipath selection.

onwards

Ignores everything as-path onwards for multipath selection.

Command Default

No default behavior or values

Command Modes

Router configuration mode

Command History

Release

Modification

Release 5.2.0

This command was introduced.

Usage Guidelines

When multiple connected routers start ignoring as-path onwards while computing multipath, it causes routing loops. Therefore, you should not configure the bgp multipath as-path ignore onwards command on routers that can form a loop.

Task ID

Task ID

Operations

bgp

read, write

Examples

This example shows how to ignore as-path while computing multipath.

RP/0/RSP0/CPU0:router# configure
RP/0/RSP0/CPU0:router(config)# router bgp 100
RP/0/RSP0/CPU0:router(config-bgp)# bgp multipath as-path ignore onwards

bgp origin-as validation enable

To enable origin-AS validation, use the bgp origin-as validation enable command, in address-family configuration mode.

bgp origin-as validation enable

This command has no arguments or keywords.

Command Default

The BGP origin-AS validation disabled.

Command Modes

Router configuration

Address family configuration

Command History

Release Modification
Release 6.5.1

This command was introduced

Usage Guidelines

None

Task ID

Task ID Operation

bgp

read, write

Examples

The following example shows how to enable origin-AS validation in IPv4 address-family configuration mode:

Router#configure
Router(config)#router bgp 50000
Router(config-bgp)#address-family ipv4 unicast
Router(config-bgp-af)#bgp origin-as validation enable

Examples

The following example shows how to enable origin-AS validation in IPv6 address-family configuration mode:

Router#configure
Router(config)#router bgp 50000
Router(config-bgp)#address-family ipv6 unicast
Router(config-bgp-af)#bgp origin-as validation enable

bgp nexthop resolution allow-default

By default, the next hop resolution in BGP does not take the default route into account. By configuring this command, the default route is used for resolving the next-hop of BGP routes. The next hop resolution is important in deciding if the next hop for a BGP route is accessible or not.

If the BGP route has an inaccessible next hop, the route does not have a best path and will not be advertised.


Warning


Applying or removing this command will reset all BGP neighbors.


bgp nexthop resolution allow-default

Syntax Description

allow-default

Enable nexthops resolution using default route.

Command Default

This applies to IPv4 and IPv6. The default route is 0.0.0.0/0 for IPv4 and ::/0 for IPv6.

Command Modes

XR Config mode

Command History

Release Modification
Release 6.2

This command was introduced.

Task ID

Task ID Operation
bgp

read, write

Examples

The following example shows how to configure BGP with nexthop resolution allow-default :

RP/0/0/CPU0:R1(config)#router bgp 65000
RP/0/0/CPU0:R1(config-bgp)#nexthop resolution allow-default

"NEXTHOP configuration changed" is seen as the last reset reason with the show bgp neighbor command when the nexthop resolution allow-default command is applied or removed:

RP/0/0/CPU0:R1#show bgp neighbor 10.0.0.2
...
  Last reset 00:01:59, due to NEXTHOP configuration changed

RP/0/0/CPU0:R1#show bgp neighbor 2001:db8:1::2
...
  Last reset 00:02:47, due to NEXTHOP configuration changed

bgp policy propagation input flow-tag

To match packets based on an incoming source, destination IP address or action (such as redirect, drop, PBTS) and redirect it to a specific VRF, use the bgp policy propagation input flow-tag command in the interface configuration mode.

bgp policy propagation input flow-tag { destination | source}

Syntax Description

bgp policy propagation input flow-tag

Enables flow-tag policy propagation on the specified interfaces.

destination

The packets are matched based on an incoming destination IP address and redirected to a specific VRF.

source

The packets are matched based on an incoming source IP address and redirect it to a specific VRF.

Command Default

None

Command Modes

Router configuration

Interface configuration

Command History

Release Modification
Release 5.3.1

This command was introduced.

Usage Guidelines

Use this command to apply the flow-tag to a specified interface. The packets are matched based on an incoming source, destination IP address or action (such as redirect, drop, PBTS) and redirected to a specific VRF.


Note


You will not be able to enable both QPPB and flow tag feature simultaneously on an interface.

Task ID

Task ID Operation
bgp

read, write

bgp redistribute-internal

To allow the redistribution of internal Border Gateway Protocol (iBGP) routes into an Interior Gateway Protocol (IGP), such as Intermediate System-to-Intermediate System (IS-IS) or Open Shortest Path First (OSPF), use the bgp redistribute-internal command in an appropriate configuration mode. To disable the redistribution of iBGP routes into IGPs, use the no form of this command.

bgp redistribute-internal

no bgp redistribute-internal

Syntax Description

This command has no keywords or arguments.

Command Default

By default, iBGP routes are not redistributed into IGPs.

Command Modes

Router configuration

VRF configuration

Command History

Release

Modification

Release 3.7.2

This command was introduced.

Usage Guidelines

To use this command, you must be in a user group associated with a task group that includes appropriate task IDs. If the user group assignment is preventing you from using a command, contact your AAA administrator for assistance.

Use of the bgp redistribute-internal command requires the clear route * command to be issued to reinstall all BGP routes into the IP routing table.


Note


Redistributing iBGP routes into IGPs may cause routing loops to form within an autonomous system. Use this command with caution.


Task ID

Task ID

Operations

bgp

read, write

Examples

The following example shows how to redistribute iBGP routes into OSPF:

RP/0/RSP0/CPU0:router(config)#router bgp 1 
RP/0/RSP0/CPU0:router(config-bgp)# bgp redistribute-internal 
RP/0/RSP0/CPU0:router(config-bgp)# exit
RP/0/RSP0/CPU0:router(config)# router ospf area1
RP/0/RSP0/CPU0:router(config-router)# redistribute bgp 1
RP/0/RSP0/CPU0:router(config-router)# end
RP/0/RSP0/CPU0:router# clear route * 

bgp router-id

To configure a fixed router ID for a Border Gateway Protocol (BGP)-speaking router, use the bgp router-id command in an appropriate configuration mode. To disable a fixed router ID, use the no form of this command.

bgp router-id ip-address

no bgp router-id [ip-address]

Syntax Description

ip-address

IP Version 4 (IPv4) address to use as the router ID. Normally, this should be an IPv4 address assigned to the router.

Command Default

If no router ID is configured in BGP, BGP attempts to use the global router ID if one is configured and available. Otherwise, BGP uses the highest IP address configured on a loopback interface.

Command Modes

Router configuration

VRF configuration

Command History

Release

Modification

Release 3.7.2

This command was introduced.

Usage Guidelines

To use this command, you must be in a user group associated with a task group that includes appropriate task IDs. If the user group assignment is preventing you from using a command, contact your AAA administrator for assistance.

If you do not use the bgp router-id command to configure a router ID, an IP address is not configured on any loopback interface, and no global router ID is configured, BGP neighbors remain down.

For more details on router IDs, see the Routing Configuration Guide for Cisco ASR 9000 Series Routers

Task ID

Task ID

Operations

bgp

read, write

Examples

The following example shows how to configure the local router with the router ID of 192.168.70.24:


RP/0/RSP0/CPU0:router(config)# router bgp 100
RP/0/RSP0/CPU0:router(config-bgp)#bgp router-id 192.168.70.24

bgp scan-time

To configure scanning intervals of Border Gateway Protocol (BGP)-speaking networking devices, use the bgp scan-time command in an appropriate configuration mode. To restore the scanning interval to its default value, use the no form of this command.

bgp scan-time seconds

no bgp scan-time seconds

Syntax Description

seconds

Scanning interval (in seconds) of BGP routing information. Range is 5 to 3600 seconds.

Command Default

The default scanning interval is 60 seconds.

Command Modes

Router configuration

IPv4 address family configuration

IPv6 address family configuration

VPNv6 address family configuration

Command History

Release

Modification

Release 3.7.2

This command was introduced.

Usage Guidelines

To use this command, you must be in a user group associated with a task group that includes appropriate task IDs. If the user group assignment is preventing you from using a command, contact your AAA administrator for assistance.

Use the bgp scan-time command to change how frequently the software processes scanner tasks, such as conditional advertisement, dynamic MED changes, and periodic maintenance tasks.

Task ID

Task ID

Operations

bgp

read, write

Examples

The following example shows how to set the scanning interval for IPv4 unicast to 20 seconds:


RP/0/RSP0/CPU0:router(config)# router bgp 64500
RP/0/RSP0/CPU0:router(config-bgp)# address-family ipv4 unicast
RP/0/RSP0/CPU0:router(config-bgp-af)# bgp scan-time 20

bgp unsafe-ebgp-policy

To pass all routes on the eBGP neighbors, use the bgp unsafe-ebgp-policy command in router configuration mode.

bgp unsafe-ebgp-policy

Syntax Description

This command has no keywords or arguments.

Command Default

None

Command Modes

Router configuration

Command History

Release

Modification

Release 7.5.1

This command was introduced.

Usage Guidelines

Configuring the bgp unsafe-ebgp-policy command enables passing all routes on the eBGP neighbors without the need to configure route policies.

Task ID

Task ID

Operations

bgp

read, write

Examples

The following example shows how to enable the unsafe ebgp policy:

Router# configure
Router(config)# router bgp 65000
Router(config-bgp)# bgp unsafe-ebgp-policy

bgp update-delay

To set the maximum initial delay for a Border Gateway Protocol (BGP)-speaking router to send the first updates, use the bgp update-delay command in an appropriate configuration mode. To restore the initial delay to its default value, use the no form of this command.

bgp update-delay seconds [always]

nobgp update-delay [seconds] [always]

Syntax Description

seconds

Delay in seconds for the router to send the first updates. Range is 0 to 3600.

always

(Optional) Specifies that the router always wait for the update delay time, even if all neighbors have finished sending their initial updates sooner.

Command Default

120 seconds

Command Modes

Router configuration

Command History

Release

Modification

Release 3.7.2

This command was introduced.

Usage Guidelines

To use this command, you must be in a user group associated with a task group that includes appropriate task IDs. If the user group assignment is preventing you from using a command, contact your AAA administrator for assistance.

When BGP is started, it waits a specified period of time for its neighbors to establish peering sessions and to complete sending their initial updates. After all neighbors complete their initial updates, or after the update delay timer expires, the best path is calculated for each route, and the software starts sending advertisements out to its peers. This behavior improves convergence time. If the software were to advertise a route as soon as it learned it, it would have to readvertise the route each time it learned a new path that was preferred over all previously learned paths.

Use the bgp update-delay command to tune the maximum time the software waits after the first neighbor is established until it starts calculating best paths and sending out advertisements.

Task ID

Task ID

Operations

bgp

read, write

Examples

The following example shows how to set the maximum initial delay to 240 seconds:


RP/0/RSP0/CPU0:router(config)#router bgp 64530 
RP/0/RSP0/CPU0:router(config-bgp)# bgp update-delay 240

bgp write-limit

To modify the upper bounds on update message queue lengths or to enable desynchronization, use the bgp write-limit command in an appropriate configuration mode. To return the bounds to their default values and to disable desynchronization, use the no form of this command.

bgp write-limit group-limit global-limit [desynchronize]

no bgp write-limit [group-limit global-limit] [desynchronize]

Syntax Description

group-limit

Per-update group limit on the number of update messages the software queues. Range is 500 to 100000000. Group limit cannot be greater than the global limit.

global-limit

Global limit on the number of update messages the software queues. Range is 500 to 100000000.

desynchronize

(Optional) Enables desynchronization.

Command Default

group-limit : 50,000

global-limit : 250,000

Desynchronizationis off.

Command Modes

Router configuration

Command History

Release

Modification

Release 3.7.2

This command was introduced.

Usage Guidelines

To use this command, you must be in a user group associated with a task group that includes appropriate task IDs. If the user group assignment is preventing you from using a command, contact your AAA administrator for assistance.

Use the bgp write-limit command to configure both a per-update group and a global limit on the number of messages the software queues when updating peers. Increasing these limits can result in faster Border Gateway Protocol (BGP) convergence, but also may result in higher memory use during convergence. In addition, this command can be used to enable desynchronization. Desynchronization can decrease memory use and speed up convergence for the fastest neighbors if one or more neighbors in an update group process updates significantly slower than other neighbors in the same group. However, enabling desynchronization can cause a significant degradation in overall convergence time, especially if the router is experiencing high CPU utilization. For this reason, enabling desynchronization is discouraged.

Task ID

Task ID

Operations

bgp

read, write

Examples

The following example shows how to configure BGP to operate with a per-update group limit of 9000 messages and a global limit of 27,000 messages:


RP/0/RSP0/CPU0:router(config)# router bgp 65000
RP/0/RSP0/CPU0:router(config-bgp)#bgp write-limit 9000 27000 

bmp-activate

To enable Border Gateway Protocol (BGP) Monitoring Protocol (BMP) logging for a neighbor, use the bmp-activate server command in neighbor configuration mode. To disable BMP logging for a neighbor, use the no form of this command.

bmp-activate server server-id

Syntax Description

server server-id

Enables monitoring by the BMP server specified by the server-id variable. You can configure multiple bmp-activate commands under same neighbor with different server IDs to enable monitoring by multiple BMP servers.

Command Default

No default behavior or values

Command Modes

Neighbor configuration

Command History

Release

Modification

Release 5.2.2

This command was introduced.

Usage Guidelines

To use this command, you must be in a user group associated with a task group that includes appropriate task IDs. If the user group assignment is preventing you from using a command, contact your AAA administrator for assistance.

Task ID

Task ID

Operations

bgp

read

Examples

The following example shows how to activate BMP on a neighbor with IP address 1.1.1.1, which is monitored by BMP server with server ID as 4:


RP/0/RSP0/CPU0:router# configure
RP/0/RSP0/CPU0:router(config)# router bgp 100
RP/0/RSP0/CPU0:router(config-bgp)# neighbor 1.1.1.1
RP/0/RSP0/CPU0:router(config-bgp-nbr)# bmp-activate server 4

bmp server

To configure Border Gateway Protocol (BGP) Monitoring Protocol (BMP) server and to enter BMP server configuration mode, use the bmp server command in Global Configuration mode. To remove a particular BMP server configuration, use no form of this command.

bmp server server-id

Syntax Description

server-id

Specifies BMP server ID. Server ID range is 1 to 8.

description LINE

Specifies BMP server description. Description can be up to 250 alphanumeric characters.

dscp

Sets IP DiffServ CodePoint (DSCP).

The DSCP value can be a number from 0 to 63, or it can be one of the following keywords: default , ef , af11 , af12 , af13 , af21 , af22 , af23 , af31 , af32 , af33 , af41 , af42 , af43 , cs1 , cs2 , cs3 , cs4 , cs5 , cs6 , or cs7 .

host host-name

Specifies the hostname of BMP server.

The hostname of the BMP server can be specified in IP address format (standard dot-decimal notation for IPv4 or colon-hexadecimal notation for IPv6) format or the string name which can be resolved into an IP address by the router.

initial-delay delay-time

Sets the delay, in seconds, before initial connect request is sent to a BMP server. The delay that you can set ranges from 1 to 3600 seconds. The default is 7 seconds.

initial-refresh {delay | skip}

Delay to initiate route refresh requests to BMP enabled neighbors.

Configures the initial refresh options to handle refresh requests sent by the router to its BMP-enabled neighbors.

Sets the delay, in seconds, before an initial refresh request is sent by the router to its BMP-enabled neighbors. The delay range is 1 to 3600 seconds with a default of 1 second. The default is not to skip refresh requests.

Configures the router to skip sending any refresh requests to its BMP-enabled neighbors.

precedence

Sets the precedence values in the IP header. The precedence value can be a number from 0 to 7, or it can be one of the following keywords:

critical —Set packets with critical precedence (5)

flash — Set packets with flash precedence (3)

flash-override —Set packets with flash override precedence (4)

immediate —Set packets with immediate precedence (2)

internet —Set packets with internetwork control precedence (6)

network —Set packets with network control precedence (7)

priority —Set packets with priority precedence (1)

routine —Set packets with routine precedence (0)

The default is internet (6) .

shutdown

Shuts down the TCP connection to BMP server.

stats-reporting-period

Specifies statistics reporting period, in seconds, to BMP servers. The reporting period that you can set ranges from 1 to 3600 seconds.

The default is 0.

update-source type interface-path-id

Specifies the source (physical or virtual interface) to reach the BMP server.

Note

 

Use the show interfaces command to see a list of all interfaces currently configured on the router.

For more information about the syntax for the router, use the question mark ( ? ) online help function.

vrf vrf-name

Specifies VPN routing and forwarding (VRF) instance.

Command Default

For default values refer Syntax Description table.

Command Modes

Global Configuration mode

Command History

Release

Modification

Release 5.2.2

This command was introduced.

Usage Guidelines

To use this command, you must be in a user group associated with a task group that includes appropriate task IDs. If the user group assignment is preventing you from using a command, contact your AAA administrator for assistance.

Task ID

Task ID

Operations

bgp

read

Examples

This example shows how to configure initial refresh delay of 30 seconds for BGP neighbors on BMP server with server ID as 4:


RP/0/RSP0/CPU0:router# configure
RP/0/RSP0/CPU0:router(config)# bmp server 4 initial-refresh delay 30

This example shows how to configure hostname of BMP server as 192.168.10.1:

RP/0/RSP0/CPU0:router# configure
RP/0/RSP0/CPU0:router(config)# bmp server 8 host 192.168.10.1 port 56

This example shows how to configure GigabitEthernet at location 0/0/0/1 as source interface to reach BMP server:

RP/0/RSP0/CPU0:router# configure
RP/0/RSP0/CPU0:router(config)# bmp server 5 update-source gigabitEthernet 0/0/0/1

bmp advertisement-interval

To set the minimum interval between the sending of BMP routing updates, use the advertisement-interval  command in router configuration mode.

Use the no form of this command to remove the advertisement-interval  command from the configuration file and restore the system to its default interval values.

advertisement-interval seconds

Syntax Description

seconds

Minimum interval between sending BMP routing updates (in seconds).

The range of the advertisement-interval of the route-monitoring inbound post-policy and the Local-RIB is from 2 seconds to 600 seconds.

The default value is 15 seconds.

Command Default

None

Command Modes

Router configuration

Command History

Release

Modification

Release 7.5.4

This command was introduced.

Usage Guidelines

  • To use this command, you must be in a user group associated with a task group that includes appropriate task IDs. If the user group assignment is preventing you from using a command, contact your AAA administrator for assistance.

  • You can configure advertisement-interval  command for the following sub-modes to configure the interval between BMP route-monitoring updates:

    • route-monitoring inbound post-policy

    • route-monitoring local-rib

Task ID

Task ID

Operations

bgp

read, write

Examples

The following example configures an advertisement-interval for 15 seconds. This is to configure the interval between BMP route-monitoring updates for Inbound Post-Policy and Local-RIB:

Router#config 
Router(config)#bmp server all 
Router(config-bgp-bmp)#route-monitoring inbound post-policy 
Router(config-bgp-bmp-rmon)#advertisement-interval 15 
Router(config-bgp-bmp-rmon)#commit 
Router#config 
Router(config)#bmp server all 
Router(config-bgp-bmp)#route-monitoring local-rib 
Router(config-bgp-bmp-rmon)#advertisement-interval 15 
Router(config-bgp-bmp-rmon)#commit 

bmp scan-time

To configure scanning intervals of BMP-speaking networking devices, use the bmp scan-time  command in router configuration mode.

Use the no form of this command to restore the scanning interval to its default value.

scan-time seconds

Syntax Description

seconds

Scanning interval (in seconds) of BMP routing information.

The range of the scan time of route-monitoring inbound post-policy is from 5 seconds to 3600 seconds.

The default value is 60 seconds.

Command Default

None

Command Modes

Router configuration

Command History

Release

Modification

Release 7.5.4

This command was introduced.

Usage Guidelines

  • To use this command, you must be in a user group associated with a task group that includes appropriate task IDs. If the user group assignment is preventing you from using a command, contact your AAA administrator for assistance.

  • You can configure BMP scan time for the route-monitoring inbound post-policy sub-mode only.

Task ID

Task ID

Operations

bgp

read, write

Examples

The following example displays the scan-time configuration for 60 seconds.

Router##config 
Router#(config)#bmp server all 
Router#(config-bgp-bmp)#route-monitoring inbound post-policy 
Router#(config-bgp-bmp-rmon)#scan-time 60 
Router#(config-bgp-bmp-rmon)#commit 

bind-source (rpki-server)

To specify a source interface for an RPKI session,, use the bind-source command in BGP rpki-server configuration mode. To remove the source interface for an RPKI session, use the no form of this command.

bind-source interface

no bind-source interface

Syntax Description

type

Interface type. For more information, use the question mark (? ) online help function.

interface-path-id

Physical interface or virtual interface.

Note

 

Use the show interfaces command to see a list of all interfaces currently configured on the router.

For more information about the syntax for the router, use the question mark (? ) online help function.

Command Default

Bind source is not configured.

Command Modes

RPKI server configuration

Command History

Release Modification

6.6.3

This command was introduced.

Usage Guidelines

Use this command to specify the source address of the rpki session.

Task ID

Task ID Operation

bgp

read, write

Examples

This example shows how to specify a source interface for an RPKI session and remove a source interface for an RPKI session:


RP/0/RSP0/CPU0:router#configure
RP/0/RSP0/CPU0:router(config)#router bgp 65000
RP/0/RSP0/CPU0:router(config-bgp)#rpki server 192.0.2.0 
RP/0/RSP0/CPU0:router(config-bgp-rpki-server)# bind-source interface loopback 0
RP/0/RSP0/CPU0:router(config-bgp-rpki-server)#commit
RP/0/RSP0/CPU0:router(config-bgp-rpki-server)#no bind-source interface loopback 0
RP/0/RSP0/CPU0:router(config-bgp-rpki-server)#commit

capability additional-paths receive

To advertise capability of receiving additional paths to the peer, use the capability additional-paths receive command in neighbor or neighbor-group or session-group configuration mode. To disable the capability of receiving additional paths, use the no form of this command.

capability additional-paths receive [disable]

no capability additional-paths receive

Syntax Description

disable

Disables advertising capability of receiving additional paths.

Command Default

Capability is disabled.

Command Modes

Neighbor configuration

Neighbor group configuration

Session group configuration

Command History

Release Modification
Release 4.0.0

This command was introduced.

Usage Guidelines

To use this command, you must be in a user group associated with a task group that includes appropriate task IDs. If the user group assignment is preventing you from using a command, contact your AAA administrator for assistance.

Use the capability additional-paths receive command to selectively enable or disable additional paths receive capability negotiation for a particular neighbor or neighbor-group or session-group. Configuring additional-paths receive command in global address-family mode is a pre-requisite for negotiating additional paths receive capability with the peer.

If you enter the capability additional-paths receive command after some BGP sessions are established, you must restart those sessions for the new configuration to take effect. Use the clear bgp command to restart sessions.

Task ID

Task ID Operation

bgp

read, write

Examples

The following example shows how to advertise capability of receiving additional paths:


RP/0/RSP0/CPU0:router(config)#router bgp 100
RP/0/RSP0/CPU0:router(config-bgp)#neighbor 10.2.3.4 
RP/0/RSP0/CPU0:router(config-bgp-nbr)#capability additional-paths receive

capability additional-paths send

To advertise capability of sending additional paths to the peer, use the capability additional-paths send command in neighbor or neighbor-group or session-group configuration mode. To disable the capability of sending additional paths, use the no form of this command.

capability additional paths send [disable]

no capability additional paths send

Syntax Description

disable

Disables advertise additional paths send capability

Command Default

Capability is disabled.

Command Modes

Neighbor configuration

Neighbor group configuration

Session group configuration

Command History

Release Modification
Release 4.0.0

This command was introduced.

Usage Guidelines

To use this command, you must be in a user group associated with a task group that includes appropriate task IDs. If the user group assignment is preventing you from using a command, contact your AAA administrator for assistance.

Use the capability additional paths send command to selectively enable or disable additional paths send capability negotiation for a particular neighbor or neighbor-group or session-group. Configuring the additional-paths send command in global address-family mode is a pre-requisite for negotiating additional paths send capability with the peer.

You must restart the BGP sessions for the new configuration to take effect. Use the clear bgp command to restart sessions.

Task ID

Task ID Operation

bgp

read, write

Examples

The following example shows how to advertise capability of sending additional paths to the peer:


RP/0/RSP0/CPU0:router(config)# router bgp 100
RP/0/RSP0/CPU0:router(config-bgp)# neighbor 10.2.3.4 
RP/0/RSP0/CPU0:router(config-bgp-nbr)# capability additional-paths send

capability orf prefix

To advertise prefix list-based Outbound Route Filter (ORF) capability to the Border Gateway Protocol (BGP) peer, use the capability orf prefix command in an appropriate configuration mode. To remove the capability orf prefix command from the configuration file and restore the system to its default condition in which the software does not advertise the capability, use the no form of this command.

capability orf prefix {receive | send | both | none}

no capability orf prefix [receive | send | both | none]

Syntax Description

receive

Sets the capability to receive the ORF from a specified neighbor.

send

Sets the capability to send the ORF to a specified neighbor.

both

Sets the capability to receive and send the ORF from or to a specified neighbor.

none

Sets the capability to no for ORF receive or send from or to a specified neighbor.

Command Default

The routing device does not receive or send route prefix filter lists.

Command Modes

IPv4 address family group configuration

IPv6 address family group configuration

IPv4 neighbor address family configuration

VRF neighbor IPv4 address family configuration

IPv4 neighbor group address family configuration

IPv6 neighbor group address family configuration

Command History

Release

Modification

Release 3.7.2

This command was introduced.

Usage Guidelines

To use this command, you must be in a user group associated with a task group that includes appropriate task IDs. If the user group assignment is preventing you from using a command, contact your AAA administrator for assistance.

The advertisement of the prefix list ORF capability by a BGP speaker indicates whether the speaker can send prefix lists to the specified neighbor and whether it accepts prefix lists from the neighbor. The speaker sends a prefix list if it indicated the ability to send them, and if the neighbor indicated it was willing to accept them. Similarly, the neighbor sends a prefix list to the speaker if it indicated the ability to send them and the speaker indicated the willingness to accept them.


Note


The capability orf and prefix list filter specified by orf route-policy must be explicitly configured.


If the neighbor sends a prefix list and the speaker accepts it, the speaker applies the received prefix list, plus any locally configured outbound filters, to limit its outbound routing updates to the neighbor. Increased filtering prevents unwanted routing updates between neighbors and reduces resource requirements for routing update generation and processing.

Use the capability orf prefix command to set whether to advertise send and receive capabilities to the specified neighbor.


Note


Sending a receive capability can adversely affect performance, because updates sent to that neighbor cannot be replicated for any other neighbors.


If this command is configured for a neighbor group or neighbor address family group, all neighbors using the group inherit the configuration. Values of commands configured specifically for a neighbor override inherited values.

Task ID

Task ID

Operations

bgp

read, write

Examples

The following example shows how to configure the capability orf prefix command:


RP/0/RSP0/CPU0:router# configure
RP/0/RSP0/CPU0:router(config)# route-policy orfqq 
RP/0/RSP0/CPU0:router:(config-rpl)# if orf prefix in (10.0.0.0/8 ge 20) then 
RP/0/RSP0/CPU0:router(config-rpl)# pass 
RP/0/RSP0/CPU0:router(config-rpl)# endif 
RP/0/RSP0/CPU0:router(config-rpl)# if orf prefix in (1910::16 ge 120) then 
RP/0/RSP0/CPU0:router(config-rpl)# pass 
RP/0/RSP0/CPU0:router(config-rpl)# endif 
RP/0/RSP0/CPU0:router(config-rpl)# end-policy 
RP/0/RSP0/CPU0:router(config)# router bgp 65530
RP/0/RSP0/CPU0:router(config-bgp)# neighbor 10.0.101.1
RP/0/RSP0/CPU0:router(config-bgp-nbr)# remote-as 65534
RP/0/RSP0/CPU0:router(config-bgp-nbr)# address-family ipv4 unicast
RP/0/RSP0/CPU0:router(config-bgp-nbr-af)# route-policy pass-all out 
RP/0/RSP0/CPU0:router(config-bgp-nbr-af)# capability orf prefix both
RP/0/RSP0/CPU0:router(config-bgp-nbr-af)# orf route-policy orfqq 

capability suppress

To suppress capabilities from being advertised to the BGP peer, use the capability suppress command in the appropriate configuration mode. To remove the capability suppress command from the configuration and restore the system to the default condition, in which the software advertises the capability, use the no form of this command.

capability suppress {4-byte-as | all | extended-nexthop-encoding} [inheritance-disable]

Syntax Description

4-byte-as

Suppresses advertisement of the 4-byte AS capability to BGP peers.

all

Suppresses advertisement of all capabilities to BGP peers.

inheritance-disable

Prevents capability suppress being inherited from the parent.

Command Default

4-byte-as capability is advertised to the BGP peer.

Command Modes

Neighbor configuration

Neighbor group configuration

Session group configuration

Command History

Release

Modification

Release 3.7.2

This command was introduced.

Release 3.9.0

The disable keyword was replaced with the inheritance-disable keyword.

Release 6.6.3

The extended-nexthop-encoding keyword was introduced.

Usage Guidelines

By default, the software advertises the 4-byte AS capability to BGP peers. To override this default behavior, use the capability suppress 4-byte-as command under the command modes listed in the ''Command Modes'' section. If configured under the neighbor group or session group, all neighbors using the group inherit the configuration. Use the no option to remove the command.


Caution


The BGP session resets automatically, if the 4-byte AS capability of an existing BGP session is changed by configuring capability suppress 4-byte-as or capability suppress 4-byte-as inheritance-disable .

Task ID

Task ID

Operations

bgp

read, write

Examples

The following example shows how to configure the capability suppress 4-byte-as command:


RP/0/RSP0/CPU0:router# show bgp nei 10.3.3.3 conf
neighbor 10.3.3.3
 remote-as 65000             [n:internal]
 description PE3             []
 update-source Loopback0     [n:internal]
 address-family ipv4 unicast [n:internal]

RP/0/RSP0/CPU0:router#show bgp nei 10.3.3.3  
BGP neighbor is 10.3.3.3
 Remote AS 65000, local AS 65000, internal link
 Description: PE3
 Remote router ID 10.3.3.3
  BGP state = Established, up for 1w0d
  Last read 00:00:17, hold time is 180, keepalive interval is 60 seconds
  Precedence: internet
  Neighbor capabilities:
    Route refresh: advertised and received
    4-byte AS: advertised and received
    Address family IPv4 Unicast: advertised and received
  Received 25962 messages, 0 notifications, 0 in queue
  Sent 25968 messages, 1 notifications, 0 in queue
  Minimum time between advertisement runs is 0 seconds

 For Address Family: IPv4 Unicast
  BGP neighbor version 1
  Update group: 0.3
  Route refresh request: received 0, sent 0
  0 accepted prefixes, 0 are bestpaths
  Prefix advertised 0, suppressed 0, withdrawn 0, maximum limit 524288
  Threshold for warning message 75%
  An EoR was received during read-only mode

  Connections established 2; dropped 1
  Last reset 1w0d, due to BGP Notification sent: hold time expired
  Time since last notification sent to neighbor: 1w0d
  Error Code: hold time expired
  Notification data sent: None

RP/0/RSP0/CPU0:router(config)#router bgp 65000
RP/0/RSP0/CPU0:router(config-bgp)#neighbor 10.3.3.3
RP/0/RSP0/CPU0:router(config-bgp-nbr)#capability suppress 4-byte-as 
RP/0/RSP0/CPU0:router(config-bgp-nbr)#commit
RP/0/RSP0/CPU0:router(config-bgp-nbr)#end 

RP/0/RSP0/CPU0:router# show bgp nei 10.3.3.3 

BGP neighbor is 10.3.3.3
 Remote AS 65000, local AS 65000, internal link
 Description: PE3
 Remote router ID 10.3.3.3
  BGP state = Established, up for 00:00:16
  Last read 00:00:11, hold time is 180, keepalive interval is 60 seconds
  Precedence: internet
  Neighbor capabilities:
    Route refresh: advertised and received
    Address family IPv4 Unicast: advertised and received
  Capability 4-byte-as suppress is configured
  Received 25966 messages, 0 notifications, 0 in queue
  Sent 25972 messages, 1 notifications, 0 in queue
  Minimum time between advertisement runs is 0 seconds

 For Address Family: IPv4 Unicast
  BGP neighbor version 1
  Update group: 0.2
  Route refresh request: received 0, sent 0
  0 accepted prefixes, 0 are bestpaths
  Prefix advertised 0, suppressed 0, withdrawn 0, maximum limit 524288
  Threshold for warning message 75%
  An EoR was received during read-only mode

  Connections established 3; dropped 2
  Last reset 00:00:43, due to Capabilty 4-byte-as configuration changed
  Time since last notification sent to neighbor: 1w0d
  Error Code: hold time expired
  Notification data sent: None

With the inheritance-disable keyword:


RP/0/RSP0/CPU0:router(config-bgp)# neighbor 10.0.101.1
RP/0/RSP0/CPU0:router(config-bgp-nbr)# capability suppress 4-byte-as inheritance-disable

RP/0/RSP0/CPU0:router# show bgp neighbor 10.0.101.1 config
neighbor 10.0.101.1
 remote-as 1                 []
 address-family ipv4 unicast []

RP/0/RSP0/CPU0:router# show bgp neighbor 10.0.101.1 
BGP neighbor is 10.0.101.1
 Remote AS 1, local AS 100, external link
 Remote router ID 0.0.0.0
  BGP state = Idle
  Last read 00:00:00, hold time is 180, keepalive interval is 60 seconds
  Precedence: internet
  Received 0 messages, 0 notifications, 0 in queue
  Sent 0 messages, 0 notifications, 0 in queue
  Minimum time between advertisement runs is 30 seconds

clear bgp

To reset a group of Border Gateway Protocol (BGP) neighbors, use the clear bgp command in EXECEXEC mode mode.

clear bgp [ipv4 {unicast | labeled-unicast | all | tunnel } | ipv6 {unicast} | all {unicast | multicast | all | labeled-unicast | tunnel} | vpnv4 unicast | vrf {vrf-name | all} {ipv4 {unicast | labeled-unicast} | ipv6 unicast}]

Syntax Description

ipv4

(Optional) Specifies IP Version 4 address prefixes.

unicast

(Optional) Specifies unicast address prefixes.

multicast

(Optional) Specifies multicast address prefixes.

labeled-unicast

(Optional) Specifies labeled unicast address prefixes.

all

(Optional) For subaddress families, specifies prefixes for all subaddress families.

tunnel

(Optional) Specifies tunnel address prefixes.

ipv6

(Optional) Specifies IP Version 6 address prefixes.

all

(Optional) For address family, specifies prefixes for all address families.

vpnv4 unicast

(Optional) Specifies VPNv4 unicast address families.

vrf

(Optional) Specifies VPN routing and forwarding (VRF).

vrf-name

Name of a VRF.

all

(Optional) For VRF, specifies all VRFs.

ipv4 { unicast | labeled-unicast }

(Optional) For VRF, specifies IPv4 unicast and labeled-unicast address families.

ipv6 unicast

(Optional) For VRF, specifies IPv6 unicast address prefixes.

Command Default

No default behavior or values

Command Modes

EXECEXEC mode

Command History

Release

Modification

Release 3.7.2

This command was introduced.

Usage Guidelines

To use this command, you must be in a user group associated with a task group that includes appropriate task IDs. If the user group assignment is preventing you from using a command, contact your AAA administrator for assistance.

Use the clear bgp command to reset the sessions of the specified group of neighbors (hard reset); it removes the TCP connection to the neighbor, removes all routes received from the neighbor from the BGP table, and then re-establishes the session with the neighbor.

If the graceful keyword is specified, the routes from the neighbor are not removed from the BGP table immediately, but are marked as stale. After the session is re-established, any stale route that has not been received again from the neighbor is removed.

Task ID

Task ID

Operations

bgp

execute

Examples

The following example shows how to hard reset neighbor 10.0.0.1:

Router# clear bgp 10.0.0.1

cef consistency-hashing auto-recovery

To enable automatic recovery of failed ECMP links and the sessions distributed due the ECMP link failure, use the cef consistent-hashing auto-recovery command in global configuration mode.

cef consistent-hashing auto-recovery

Syntax Description

This command has no keywords or arguments.

Command Default

Failed ECMP links are not automatically recovered.

Command Modes

Global configuration

Command History

Release Modification
Release 6.5.1

The command was introduced.

Usage Guidelines

Configuring the command does not alter the current state. The command takes effect on the next link down or up events.

Task ID

Task ID Operation
ipv4

read, write

Examples

Router# configure
Router(config)# cef consistent-hashing auto-recovery 

cef fast-reroute follow bgp-pic

A 4-minutes timer is being used to prevent traffic loss when using (Border Gateway Protocol) BGP with (Fast Reroute) FRR. The timer can cause an outage of up to 4 minutes when interface or BFD flaps are observed. The FRR would try to keep the backup route longer in use, but can cause a longer outage when interface or BFD flaps occurs. In order to avoid this potential outage created by the 4-minutes timer, the command cef fast-reroute follow bgp-pic is being used to disable it.

cef fast-reroute follow bgp-pic

Syntax Description

This command has no keywords or arguments.

Command Default

4-minutes timer acting upon when the FRR change is enabled.

Command Modes

Global configuration

Command History

Release Modification

Release 6.3.2

The command was introduced.

Release 7.3.1

The command was deprecated.

Usage Guidelines

Configuring the command does not alter the current state. The command is used to disable the 4-minutes timer.

Task ID

Task ID Operation
bgp

read, write

Examples

Router# configure
Router(config)# cef fast-reroute follow bgp-pic

clear bgp current-mode

To switch from one BGP mode to another, use the clear bgp current-mode command in EXEC configuration mode.

clear bgp current-mode

Syntax Description

This command has no arguments or keywords.

Command Default

No default behavior or values

Command Modes

EXEC configuration

Command History

Release

Modification

Release 3.7.2

This command was introduced.

Usage Guidelines

To use this command, you must be in a user group associated with a task group that includes appropriate task IDs. If the user group assignment is preventing you from using a command, contact your AAA administrator for assistance.

Use the clear bgp current-mode command to switch from standalone to distributed mode, or from distributed to standalone mode. The show bgp process command indicates the current BGP mode.


Note


Switching from one mode to another causes all BGP sessions to go down.


Task ID

Task ID

Operations

bgp

execute

Examples

The following example shows the show bgp process command output before and after switching from one BGP mode to another:


RP/0/RSP0/CPU0:router# show bgp process
  
  BGP Process Information
  BGP is operating in STANDALONE mode
  Autonomous System: 3
  Router ID: 10.18.18.11
  Cluster ID: 10.18.18.11
  Fast external fallover enabled
  Neighbor logging is enabled
  Enforce first AS enabled
  Default local preference: 100
  Default keepalive: 60
  Update delay: 120
  Generic scan interval: 60
  
  Address family: IPv4 Unicast
  Dampening is not enabled
  Client reflection is enabled
  Scan interval: 60
  Main Table Version: 3
  IGP notification: IGPs notified
  RIB has converged: version 0
  
  Node                Process     Nbrs Estb Rst Upd-Rcvd Upd-Sent Nfn-Rcv Nfn-Snt
  node0_0_CPU0        Speaker        5    5  51        0        7       0       5
  
  RP/0/RSP0/CPU0:router# configure 
  RP/0/RSP0/CPU0:router(config)# router bgp 3 
  RP/0/RSP0/CPU0:router(config-bgp)# distributed speaker 1
  RP/0/RSP0/CPU0:router(config-bgp)# distributed speaker 2
  RP/0/RSP0/CPU0:router(config-bgp)# neighbor 10.0.101.1
  RP/0/RSP0/CPU0:router(config-bgp-nbr)# speaker-id 2 
  RP/0/RSP0/CPU0:router(config-bgp-nbr)# commit 
  RP/0/RSP0/CPU0:router# clear bgp current-mode 
  RP/0/RSP0/CPU0:router# show bgp process 
  
  BGP Process Information
  BGP is operating in DISTRIBUTED mode
  Autonomous System: 3
  Router ID: 10.18.18.11
  Cluster ID: 10.18.18.11
  Fast external fallover enabled
  Neighbor logging is enabled
  Enforce first AS enabled
  Default local preference: 100
  Default keepalive: 60
  Update delay: 120
  Generic scan interval: 60
  
  Address family: IPv4 Unicast
  Dampening is not enabled
  Client reflection is enabled
  Scan interval: 60
  Main Table Version: 1
  IGP notification: IGPs not notified
  RIB has not converged: version 0
  
  Node                Process     Nbrs Estb Rst Upd-Rcvd Upd-Sent Nfn-Rcv Nfn-Snt
  node0_0_CPU0        Speaker 1      4    1  52        0        0       0       4
  node0_0_CPU0        Speaker 2      1    0   9        0        0       0       1
  node0_0_CPU0        bRIB 1         0    0   0        0        0       0       0
  node0_0_CPU0        bRIB 2         0    0   0        0        0       0       0
  

clear bgp dampening

To clear Border Gateway Protocol (BGP) route dampening information and unsuppress the suppressed routes, use the clear bgp dampening command in EXEC configurationEXEC mode mode.

clear bgp dampening

Syntax Description

ipv4

Specifies IP Version 4 address prefixes.

unicast

Specifies unicast address prefixes.

multicast

Specifies multicast address prefixes.

all

For subaddress families, specifies prefixes for all subaddress families.

ipv6

Specifies IP Version 6 address prefixes.

all

For address family, specifies prefixes for all address families.

vpnv4 unicast

Specifies VPNv4 unicast address families.

vrf

Specifies VPN routing and forwarding (VRF).

vrf-name

Name of a VRF.

all

For VRF, specifies all VRFs.

ipv4 { unicast | labeled-unicast }

For VRF, specifies IPv4 unicast and labeled-unicast address families.

ipv6 unicast

For VRF, specifies IPv6 unicast address families.

ip-address

(Optional) IP address of the network about which to clear dampening information.

/mask-length

(Optional) Network mask applied to the IP address.

Command Default

If no IP address is specified, dampening information for all routes is cleared.

Command Modes

EXEC configurationEXEC mode

Command History

Release

Modification

Release 3.7.2

This command was introduced.

Usage Guidelines

To use this command, you must be in a user group associated with a task group that includes appropriate task IDs. If the user group assignment is preventing you from using a command, contact your AAA administrator for assistance.

Always use the clear bgp dampening command for an individual address-family. The all option for address-families with clear bgp dampening should never be used during normal functioning of the system. For example, use
clear bgp ipv4 unicast dampening prefix x.x.x./y

Task ID

Task ID

Operations

bgp

execute

Examples

The following example shows how to clear the route dampening information for all 172.20.0.0/16 IPv4 unicast paths:


Router# clear bgp ipv4 unicast dampening 172.20.0.0/16

clear bgp external

To clear all Border Gateway Protocol (BGP) external peers, use the clear bgp external command in EXEC configuration mode.

clear bgp external

Syntax Description

ipv4

(Optional) Specifies IP Version 4 address prefixes.

unicast

(Optional) Specifies unicast address prefixes.

multicast

(Optional) Specifies multicast address prefixes.

labeled-unicast

(Optional) Specifies labeled unicast address prefixes.

all

(Optional) For subaddress families, specifies prefixes for all subaddress families.

ipv6

(Optional) Specifies IP Version 6 address prefixes.

all

(Optional) For address family, specifies prefixes for all address families.

vpnv4 unicast

(Optional) Specifies VPNv4 unicast address families.

vrf

(Optional) Specifies VPN routing and forwarding (VRF).

vrf-name

(Optional) Name of a VRF.

all

(Optional) For VRF, specifies all VRFs.

ipv4 { unicast | labeled-unicast }

(Optional) For VRF, specifies IPv4 unicast or labeled-unicast address families.

ipv6 unicast

(Optional) For VRF, specifies IPv6 unicast address families.

vpnv6 unicast

(Optional) Specifies VPNv6 unicast address families.

graceful

(Optional) Clears all external peers with a hard reset and a graceful restart. This option is available when an address family is not specified.

Command Default

No default behavior or value

Command Modes

EXEC configuration

Command History

Release

Modification

Release 3.7.2

This command was introduced.

Usage Guidelines

To use this command, you must be in a user group associated with a task group that includes appropriate task IDs. If the user group assignment is preventing you from using a command, contact your AAA administrator for assistance.

Task ID

Task ID

Operations

bgp

execute

Examples

The following example shows how to clear all BGP external peers:

RP/0/RSP0/CPU0:router# clear bgp external

clear bgp flap-statistics

To clear Border Gateway Protocol (BGP) flap counts for a specified group of routes, use the clear bgp flap-statistics command in EXEC configurationEXEC mode mode.

clear bgp flap-statistics

Syntax Description

ipv4

Specifies IP Version 4 address prefixes.

unicast

Specifies unicast address prefixes.

multicast

Specifies multicast address prefixes.

labeled-unicast

Specifies labeled unicast address prefixes.

all

For subaddress families, specifies prefixes for all subaddress families.

ipv6

Specifies IP Version 6 address prefixes.

all

For address family, specifies prefixes for all address families.

vpnv4 unicast

Specifies VPNv4 unicast address families.

vrf

Specifies VPN routing and forwarding (VRF).

vrf-name

Name of a VRF.

all

For VRF, specifies all VRFs.

ipv4 { unicast | labeled-unicast }

For VRF, specifies IPv4 unicast or labeled-unicast address families.

ipv6 unicast

For VRF, specifies IPv6 unicast address families.

vpnv6 unicast

Specifies VPNv6 unicast address families.

regexp regexp

(Optional) Clears flap statistics for routes whose AS paths match the regular expression.

route-policy route-policy-name

(Optional) Clears flap statistics for the specific route policy.

network

(Optional) Network for which flap counts are to be cleared.

/mask-length

(Optional) Network mask of the network for which flap counts are to be cleared.

ip-address

(Optional) Neighbor address. Clears only flap statistics for routes received from this neighbor.

Command Default

No default behavior or value

Command Modes

EXEC

Command History

Release

Modification

Release 3.7.2

This command was introduced.

Usage Guidelines

To use this command, you must be in a user group associated with a task group that includes appropriate task IDs. If the user group assignment is preventing you from using a command, contact your AAA administrator for assistance.

Task ID

Task ID

Operations

bgp

execute

Examples

The following example shows how to clear the flap count for all routes (in all address families) originating in autonomous system 1:

Router#clear bgp all all flap-statistics regexp _1$ 

The following example shows how to clear the flap count for all IPv4 unicast routes received from neighbor 172.20.1.1:

Router# clear bgp ipv4 unicast flap-statistics 172.20.1.1

clear bgp long-lived-stale

To delete all paths received from the given neighbor that are long-lived-stale, use the clear bgp long-lived-stale command in EXEC mode.

clear bgp vrf {vrf-name | all} {ipv4 | ipv6} unicast nbr-address long-lived-stale

Syntax Description

vrf vrf-name

Deletes all paths received from the given neighbor that are long-lived-stale for the specified VRF

vrf all

Deletes all paths received from the given neighbor that are long-lived-stale for all VRFs.

ipv4 unicast

Specifies IP Version 4 unicast address prefixes.

ipv6 unicast

Specifies IP Version 6 unicast address prefixes.

nbr-address

Specifies IPv4 or IPv6 address of the neighbor.

Command Default

No default behavior

Command Modes

EXEC

Command History

Release Modification
Release 5.2.2

This command was introduced

Usage Guidelines

To use this command, you must be in a user group associated with a task group that includes appropriate task IDs. If the user group assignment is preventing you from using a command, contact your AAA administrator for assistance.

Task ID

Task ID Operation
bgp

read, write

Examples

This command deletes all paths received from the given neighbor for all VRFs:

RP/0/0/CPU0:router# clear bgp vrf all ipv4 unicast 192.172.20.10 long-lived-stale

clear bgp nexthop performance-statistics

To reset the number of received notifications and the cumulative processing time for the Border Gateway Protocol (BGP) next hop, use the clear bgp nexthop performance-statistics command in EXEC configuration mode.

clear bgp nexthop performance-statistics

Syntax Description

ipv4

Specifies IP Version 4 address prefixes.

unicast

Specifies unicast address prefixes.

multicast

Specifies multicast address prefixes.

labeled-unicast

Specifies labeled unicast address prefixes.

all

For subaddress families, specifies prefixes for all subaddress families.

tunnel

Specifies tunnel address prefixes.

ipv6

Specifies IP Version 6 address prefixes.

all

For address family, specifies prefixes for all address families.

vpnv4 unicast

Specifies VPNv4 unicast address families.

vrf

Specifies VPN routing and forwarding (VRF).

vrf-name

Name of a VRF.

all

For VRF, specifies all VRFs.

ipv4 { unicast | labeled-unicast }

For VRF, specifies IPv4 unicast or labeled-unicast address families.

ipv6 unicast

For VRF, specifies IPv6 unicast address families.

Command Default

No default behavior or values

Command Modes

EXEC configuration

Command History

Release

Modification

Release 3.7.2

This command was introduced.

Usage Guidelines

To use this command, you must be in a user group associated with a task group that includes appropriate task IDs. If the user group assignment is preventing you from using a command, contact your AAA administrator for assistance.

Use the clear bgp nexthop performance-statistics command to reset the total number of notifications received from the Routing Information Base (RIB) and the cumulative next-hop processing time. The following information is cleared from the show bgp nexthops command output:

  • Total critical notifications received
  • Total noncritical notifications received
  • Best path deleted after last walk
  • Best path changed after last walk
  • Next-hop table total number of critical and noncritical notifications (Notf) and the time of the last notification received from the RIB (LastRIB) columns (only entries that have a status of unreachable [UR])

Task ID

Task ID

Operations

bgp

execute

Examples

The following example shows how to clear next-hop performance statistics:

RP/0/RSP0/CPU0:router# clear bgp vrf vrf_A nexthop performance statistics 

clear bgp nexthop registration

To reregister a specified next hop with the Routing Information Base (RIB), use the clear bgp nexthop registration command in EXEC configuration mode.

clear bgp nexthop registration nexthop-address nexthop-address

Syntax Description

ipv4

Specifies IP Version 4 address prefixes.

unicast

Specifies unicast address prefixes.

multicast

Specifies multicast address prefixes.

labeled-unicast

Specifies labeled-unicast address prefixes.

all

For subaddress families, specifies prefixes for all subaddress families.

tunnel

Specifies tunnel address prefixes.

ipv6

Specifies IP Version 6 address prefixes.

all

For address family, specifies prefixes for all address families.

vpnv4 unicast

Specifies VPNv4 unicast address families.

vrf

Specifies VPN routing and forwarding (VRF).

vrf-name

Name of a VRF.

all

For VRF, specifies all VRFs.

ipv4 { unicast | labeled-unicast }

For VRF, specifies IPv4 unicast or labeled-unicast address families.

ipv6 unicast

For VRF, specifies IPv6 unicast address families.

nexthop-address

Address of the next hop.

Command Default

No default behavior or values

Command Modes

EXEC configuration

Command History

Release

Modification

Release 3.7.2

This command was introduced.

Usage Guidelines

To use this command, you must be in a user group associated with a task group that includes appropriate task IDs. If the user group assignment is preventing you from using a command, contact your AAA administrator for assistance.

Use the clear bgp nexthop registration command to perform an asynchronous registration of the next hop with the RIB. The show bgp nexthops command output shows a critical notification as the LastRIBEvent for the next hop when the clear bgp nexthop registration command is used.

Task ID

Task ID

Operations

bgp

execute

Examples

The following example shows how to reregister the next hop with the RIB:

RP/0/RSP0/CPU0:router# clear bgp nexthop registration 10.1.1.1 

clear bgp peer-drops

To clear the connection-dropped counter, use the clear bgp peer-drops command in EXEC configuration mode.

clear bgp peer-drops {* | ip-address}

Syntax Description

*

Specifies all BGP neighbors.

ip-address

IP address of a specific network neighbor.

Command Default

No default behavior or values

Command Modes

EXEC configuration

Command History

Release

Modification

Release 3.7.2

This command was introduced.

Usage Guidelines

To use this command, you must be in a user group associated with a task group that includes appropriate task IDs. If the user group assignment is preventing you from using a command, contact your AAA administrator for assistance.

Task ID

Task ID

Operations

bgp

execute

Examples

The following example shows how to clear the connection-dropped counter for all BGP neighbors:

RP/0/RSP0/CPU0:router# clear bgp peer-drops *

clear bgp performance-statistics

To clear the performance statistics for all address families, use the clear bgp performance-statistics command.

clear bgp [vrf {vrf-name | all}] performance-statistics

Syntax Description

vrf

Specifies VPN routing and forwarding (VRF).

vrf-name

Name of a VRF.

all

For VRF, specifies all VRFs.

Command Default

No default behavior or values

Command Modes

EXEC configuration

Command History

Release

Modification

Release 3.7.2

This command was introduced.

Usage Guidelines

To use this command, you must be in a user group associated with a task group that includes appropriate task IDs. If the user group assignment is preventing you from using a command, contact your AAA administrator for assistance.

Task ID

Task ID

Operations

bgp

execute

Examples

The following example shows how to clear the performance statistics for all address families:

RP/0/RSP0/CPU0:router# clear bgp performance-statistics

clear bgp rpki

To clear resource public key infrastructure (RPKI) cache-server (rpki-sever) or Prefix validation state, use the clear bgp rpki command.

Use the server option in the clear bgp rpki command in EXEC mode to clear the Resource Public Key Infrastructure (RPKI) server, purge Route Origin Authorizations (ROA) records downloaded from that server, renegotiate the TCP connection, and redownload ROA records.

Use clear bgp rpki server WORD command to clear a specific RPKI cache-server (hostname or IP address).

The clear bgp rpki server * command clears all RPKI cache-servers.

Use the clear bgp rpki validation command to re-run validation for all prefixes under the specified VRF or all VRFs, if not specified. If the address family is specified after the command, then only prefixes belonging to that address family will have their validation checks re-run.

clear bgp rpki server { * | WORD }

clear bgp { vrf <vrf name> } rpki validation { ipv4 | ipv6 }

Syntax Description

server

Clear RPKI cache-server.

validation

Clear RPKI origin-validation (trigger origin validation).

*

Clear all RPKI cache-servers.

WORD

Clear a specific RPKI cache-server (hostname or IP address).

ipv4

Trigger origin validation walk for the IPv4 table.

ipv6

Trigger origin validation walk for the IPv6 table.

Command Modes

EXEC

Command History

Release Modification

Release 4.2.1

This command was introduced.

Examples

The following example shows how to configure BGP RPKI server:

router bgp 65000
 rpki server 10.0.101.2
  bind-source interface GigabitEthernet0/2/0/1
  transport tcp port 32001
  response-time 15
The following example shows the output of show bgp rpki server summary and show bgp rpki table commands:

RP/0/0/CPU0:router#show bgp rpki server summary 
Wed May  6 10:26:42.532 PDT

Hostname/Address        Transport       State           Time            ROAs (IPv4/IPv6)
10.0.101.2              TCP:32001       ESTAB           00:00:09        3/3

RP/0/0/CPU0:router#show bgp rpki table        
Wed May  6 10:26:44.592 PDT

  Network               Maxlen          Origin-AS         Server
  1.1.1.0/24            32              9991              10.0.101.2
  67.64.0.0/30          32              4002              10.0.101.2
  67.128.0.0/30         32              4003              10.0.101.2

  Processed 3 RPKI entries
The following example shows the output of show bgp rpki server summary command after clearing RPKI server 10.0.101.2:

RP/0/0/CPU0:router#clear bgp rpki server 10.0.101.2
Wed May  6 10:26:50.318 PDT

RP/0/0/CPU0:router#show bgp rpki server summary    
Wed May  6 10:26:53.212 PDT

Hostname/Address        Transport       State           Time            ROAs (IPv4/IPv6)
10.0.101.2              TCP:32001       NONE            00:00:04        N/A

RP/0/0/CPU0:router#show bgp rpki table             
Wed May  6 10:26:55.837 PDT

  Processed 0 RPKI entries

RP/0/0/CPU0:router#show bgp rpki server summary
Wed May  6 10:27:04.663 PDT

Hostname/Address        Transport       State           Time            ROAs (IPv4/IPv6)
10.0.101.2              TCP:32001       ESTAB           00:00:11        0/0

RP/0/0/CPU0:router#show bgp rpki table 
Wed May  6 10:27:17.513 PDT

  Network               Maxlen          Origin-AS         Server
  1.1.1.0/24            32              9991              10.0.101.2
  67.64.0.0/30          32              4002              10.0.101.2
  67.128.0.0/30         32              4003              10.0.101.2

  Processed 3 RPKI entries

clear bgp self-originated

To clear Border Gateway Protocol (BGP) routes that are self-originated, use the clear bgp self-originated command in EXEC configurationEXEC mode mode.

clear bgp {ipv4 {unicast | | multicast | | labeled-unicast | | all} | | ipv6 {unicast | | multicast | | labeled-unicast | | all | } | | all {unicast | | multicast | | labeled-unicast | | all} | | vpnv4 unicast | | vrf {vrf-name | | all} | | vpnv6 unicast} self-originated

Syntax Description

ipv4

Specifies IP Version 4 address prefixes.

unicast

Specifies unicast address prefixes.

multicast

Specifies multicast address prefixes.

labeled-unicast

Specifies labeled unicast address prefixes.

all

For subaddress families, specifies prefixes for all subaddress families.

ipv6

Specifies IP Version 6 address prefixes.

all

For address family, specifies prefixes for all address families.

vpnv4 unicast

Specifies VPNv4 unicast address families.

vrf

Specifies VPN routing and forwarding (VRF).

vrf-name

Name of a VRF.

all

For VRF, specifies all VRFs.

ipv4 { unicast | labeled-unicast }

For VRF, specifies IPv4 unicast or labeled-unicast address families.

ipv6 unicast

For VRF, specifies IPv6 unicast address families.

Command Default

No default behavior or values

Command Modes

EXEC configurationEXEC mode

Command History

Release

Modification

Release 3.7.2

This command was introduced.

Usage Guidelines

To use this command, you must be in a user group associated with a task group that includes appropriate task IDs. If the user group assignment is preventing you from using a command, contact your AAA administrator for assistance.

Self-originated routes are routes locally originated by the network command, redistribute command, or aggregate-address command.

Task ID

Task ID

Operations

bgp

execute

Examples

The following example shows how to clear self-originated IPv4 routes:


Router# clear bgp ipv4 unicast self-originated

clear bgp shutdown

To clear all Border Gateway Protocol (BGP) neighbors that shut down due to low memory, use the clear bgp shutdown command in EXEC configuration mode.

clear bgp {ipv4 {unicast | | multicast | | labeled-unicast | | all} | | ipv6 {unicast | | multicast | | labeled-unicast | | all | } | | all {unicast | | multicast | | labeled-unicast | | all} | | vpnv4 unicast | | vrf {vrf-name | | all} | | vpnv6 unicast} shutdown

Syntax Description

ipv4

Specifies IP Version 4 address prefixes.

unicast

Specifies unicast address prefixes.

multicast

Specifies multicast address prefixes.

labeled-unicast

Specifies labeled unicast address prefixes.

all

For subaddress families, specifies prefixes for all subaddress families.

ipv6

Specifies IP Version 6 address prefixes.

all

For address family, specifies prefixes for all address families.

vpnv4 unicast

Specifies VPNv4 unicast address families.

vrf

Specifies VPN routing and forwarding (VRF).

vrf-name

Name of a VRF.

all

For VRF, specifies all VRFs.

ipv4 { unicast | labeled-unicast }

For VRF, specifies IPv4 unicast or labeled-unicast address families.

Command Default

No default behavior or values

Command Modes

EXEC configuration

Command History

Release

Modification

Release 3.7.2

This command was introduced.

Usage Guidelines

To use this command, you must be in a user group associated with a task group that includes appropriate task IDs. If the user group assignment is preventing you from using a command, contact your AAA administrator for assistance.

Task ID

Task ID

Operations

bgp

execute

Examples

The following example shows how to clear all shut-down BGP neighbors:


RP/0/RSP0/CPU0:router# clear bgp shutdown

clear bgp soft

To soft reset a group of Border Gateway Protocol (BGP) neighbors, use the clear bgp soft command in EXEC configurationEXEC mode mode.

clear bgp {ipv4 {unicast | | multicast | | labeled-unicast | | all | | tunnel | | mdt} | | ipv6 {unicast | | multicast | | labeled-unicast | | all | } | | all {unicast | | multicast | | labeled-unicast | | all | | tunnel | | mdt} | | vpnv4 unicast | | vrf {vrf-name | | all} | | vpnv6 unicast} {* | | ip-address | | as | as-number | | external}soft [ [in | {prefix-filter} | | out]]

Syntax Description

ipv4

Specifies IP Version 4 address prefixes.

unicast

Specifies unicast address prefixes.

multicast

Specifies multicast address prefixes.

labeled-unicast

Specifies labeled unicast address prefixes.

all

For subaddress families, specifies prefixes for all subaddress families.

tunnel

Specifies tunnel address prefixes.

ipv6

Specifies IP Version 6 address prefixes.

all

For address family, specifies prefixes for all address families.

vpnv4 unicast

Specifies VPNv4 unicast address families.

vrf

Specifies VPN routing and forwarding (VRF).

vrf-name

Name of a VRF.

all

For VRF, specifies all VRFs.

ipv4 { unicast | labeled-unicast }

For VRF, specifies IPv4 unicast or labeled-unicast address families.

ipv6 unicast

For VRF, specifies IPv6 unicast address families.

*

Soft resets all BGP neighbors.

ip-address

IP address of the neighbor to be reset.

as as-number

Autonomous system (AS) number for all neighbors to be reset. Range for 2-byte numbers is 1 to 65535. Range for 4-byte numbers is 1.0 to 65535.65535.

external

Specifies clearing of all external peers.

in

(Optional) Triggers an inbound soft reset. If the in or out keyword is not specified, both inbound and outbound soft resets are triggered.

prefix-filter

(Optional) Specifies to send a new Outbound Route Filter (ORF) to the neighbor. Neighbor installs the new ORF and resends its routes.

out

(Optional) Triggers an outbound soft reset. If the in or out keyword is not specified, both inbound and outbound soft resets are triggered.

Command Default

No default behavior or value

Command Modes

EXEC configurationEXEC mode

Command History

Release

Modification

Release 3.7.2

This command was introduced.

Usage Guidelines

To use this command, you must be in a user group associated with a task group that includes appropriate task IDs. If the user group assignment is preventing you from using a command, contact your AAA administrator for assistance.

Use the clear bgp soft command to trigger a soft reset of the specified address families for the specified group of neighbors. This command is useful if you change the inbound or outbound policy for the neighbors, or any other configuration that affects the sending or receiving of routing updates.

If an outbound soft reset is triggered, BGP resends all routes for the address family to the given neighbors.

If an inbound soft reset is triggered, BGP by default sends a REFRESH request to the neighbor, if the neighbor has advertised the ROUTE_REFRESH capability. To determine whether the neighbor has advertised the ROUTE_REFRESH capability, use the show bgp neighbors command, and look for the following line of output:


  Received route refresh capability from peer.
  
  

If the neighbor does not support route refresh, but the soft-reconfiguration inbound command is configured for the neighbor, then BGP uses the routes cached as a result of the soft-reconfiguration inbound command to perform the soft reset.

If you want BGP to use the cached routes even if the neighbor supports route refresh, you can use the always keyword when configuring the soft-reconfiguration inbound command.

If the neighbor does not support route refresh and the soft-reconfiguration inbound command is not configured, then inbound soft reset is not possible. In this case, an error is printed.


Note


By default, if the configuration for an inbound or outbound route policy is changed, BGP performs an automatic soft reset. Use the bgp auto-policy-soft-reset disable command to disable this behavior.


Task ID

Task ID

Operations

bgp

execute

Examples

The following example shows how to trigger an inbound soft clear for IPv4 unicast routes received from neighbor 10.0.0.1:


RP/0/RP0/CPU0:router# clear bgp ipv4 unicast 10.0.0.1 soft in

cluster-id allow-equal

To enable a router to accept BGP routes which have the same first cluster-ID as the router’s own cluster-ID in the list of cluster-IDs, use the cluster-id allow-equal command in BGP neighbor address-family configuration mode. To revert to the default behaviour, use the disable option, or no form of this command.

cluster-id allow-equal [disable]

Syntax Description

disable

Specifies that BGP routes which have the same first cluster-ID as the router’s own cluster-ID in the list of cluster-IDs, is rejected.

Command Default

A router rejects BGP routes which have the same first cluster-ID as the router’s own cluster-ID in the list of cluster-IDs.

Command Modes

BGP neighbor address-family configuration.

Command History

Release

Modification

Release 6.6.3

This command was introduced.

Usage Guidelines

To avoid routing loops, the cluster-id allow-equal configuration must be done with care.

Examples

The following example shows how to enable a router to accept BGP routes which have the same first cluster-ID as the router’s own cluster-ID in the list of cluster-IDs.


Router(config)# router bgp 100
Router(config-bgp)# neighbor 10.1.1.1
Router(config-bgp-nbr)# address-family ipv4 unicast  
Router(config-bgp-nbr-af)# cluster-id allow-equal

The following example shows how to disable a router from accepting BGP routes from the same cluster-ID as the router’s own cluster-ID in the list of cluster-IDs.


Router(config)# router bgp 100
Router(config-bgp)# neighbor 10.1.1.1
Router(config-bgp-nbr)# address-family ipv4 unicast  
Router(config-bgp-nbr-af)# cluster-id allow-equal disable

default-information originate (BGP)

To allow origination of a default route to be redistributed into the Border Gateway Protocol (BGP) from another protocol, use the default-information originate command in an appropriate configuration mode. To disable this function, use the no form of this command.

default-information originate

no default-information originate

Syntax Description

This command has no arguments or keywords.

Command Default

BGP does not permit redistribution of a default route into BGP.

Command Modes

Router configuration

VRF configuration

Command History

Release

Modification

Release 3.7.2

This command was introduced.

Usage Guidelines

To use this command, you must be in a user group associated with a task group that includes appropriate task IDs. If the user group assignment is preventing you from using a command, contact your AAA administrator for assistance.

Use the redistribute command to redistribute routes from another protocol into BGP. By default, if these routes include the default route (0.0.0.0/0 for IPv4 or ::/0 for IPv6), the default route is ignored. Use the default-information originate command to change this behavior so that the default route is not ignored and is redistributed into BGP along with the other routes for the protocol being redistributed.

Task ID

Task ID

Operations

bgp

read, write

Examples

The following example shows how to configure BGP to redistribute the default route into BGP:


RP/0/RSP0/CPU0:router(config)#router bgp 164 
RP/0/RSP0/CPU0:router(config-bgp)# default-information originate

default-martian-check disable

To disable the Martian check on the IPv4 and IPv6 prefixes, use the default-martian-check disable command in the address-family configuration mode. To enable the Martian check on the IPv4 and IPv6 prefixes, use the no form of this command.

default-martian-check disable

no default-martian-check disable

Syntax Description

This command has no keywords or arguments.

Command Default

None

Command Modes

IPv4 address family configuration mode.

IPv6 address family configuration mode.

Command History

Release

Modification

Release 5.1

This command was introduced.

Usage Guidelines

To use this command, you must be in a user group associated with a task group that includes appropriate task IDs. If the user group assignment is preventing you from using a command, contact your AAA administrator for assistance.

Task ID

Task ID

Operations

bgp

read, write

Examples

This example shows how to disable Martian check for an IPv4 address prefix.


RP/0/RSP0/CPU0:router# configure
RP/0/RSP0/CPU0:router(config)# router bgp 100
RP/0/RSP0/CPU0:router(config-bgp)# address-family ipv4 multicast
RP/0/RSP0/CPU0:router(config-bgp-af)# default-martian-check disable  

This example shows how to disable Martian check for an IPv6 address prefix.


RP/0/RSP0/CPU0:router# configure
RP/0/RSP0/CPU0:router(config)# router bgp 100
RP/0/RSP0/CPU0:router(config-bgp)# address-family ipv6 multicast
RP/0/RSP0/CPU0:router(config-bgp-af)# default-martian-check disable  

default-metric (BGP)

To set default metric values for the Border Gateway Protocol (BGP), use the default-metric command in an appropriate configuration mode. To disable metric values, use the no form of this command.

default-metric value

no default-metric [value]

Syntax Description

value

Default metric value appropriate for the specified routing protocol. Range is 1 to 4294967295.

Command Default

A metric is not set.

Command Modes

Router configuration

VRF configuration

Command History

Release

Modification

Release 3.7.2

This command was introduced.

Usage Guidelines

To use this command, you must be in a user group associated with a task group that includes appropriate task IDs. If the user group assignment is preventing you from using a command, contact your AAA administrator for assistance.

Use the default-metric command to set the Multi Exit Discriminator (MED) to advertise to peers for routes that do not already have a metric set (routes that were received with no MED attribute).


Note


The metric values that you apply using the default-metric command take effect only for a new prefix which gets into the BGP table. The metrics for the existing prefixes in the BGP table remain the same. Also, when you remove the default-metric command from the configuration, the metrics which were previously assigned for prefixes are not updated. To get out of this condition, clear the BGP neighborship.


Task ID

Task ID

Operations

bgp

read, write

Examples

The following example shows how to set the BGP default metric:


RP/0/RSP0/CPU0:router(config)# router bgp 109
RP/0/RSP0/CPU0:router(config-bgp)# default-metric 10

default-originate

To cause a Border Gateway Protocol (BGP) speaker (the local router) to send the default route 0.0.0.0/0 to a neighbor for use as a default route, use the default-originate command in an appropriate configuration mode. To disable this function, use the no form of this command.

default-originate [inheritance-disable | route-policy route-policy-name]

no default-originate [inheritance-disable | route-policy route-policy-name]

Syntax Description

inheritance-disable

(Optional) Prevents the default-originate command characteristics from being inherited from a parent group.

route-policy route-policy-name

(Optional) Specifies the name of a route policy. The route policy allows route 0.0.0.0 to be injected conditionally. IPv6 address family is supported.

Command Default

The default route is not advertised to BGP neighbors.

Command Modes

IPv4 neighbor address family configuration

IPv6 neighbor address family configuration

IPv4 neighbor group address family configuration

IPv6 neighbor group address family configuration

IPv4 address family group configuration

IPv6 address family group configuration

VRF IPv4 neighbor address family configuration

VRF IPv6 neighbor address family configuration

Command History

Release

Modification

Release 3.7.2

This command was introduced.

Release 3.9.0

The disable keyword was replaced with the inheritance-disable keyword.

Usage Guidelines

To use this command, you must be in a user group associated with a task group that includes appropriate task IDs. If the user group assignment is preventing you from using a command, contact your AAA administrator for assistance.

The default-originate command does not require the presence of the default route (0.0.0.0/0 for IPv4 or ::/0 for IPv6) in the local router. When the default-originate command is used with a route policy, the default route is advertised if any route in the BGP table matches the policy.

Task ID

Task ID

Operations

bgp

read, write

Examples

The following example shows how to unconditionally advertise the route 0.0.0.0/0 to the neighbor 172.20.2.3:

RP/0/RSP0/CPU0:router(config)# router bgp 109
RP/0/RSP0/CPU0:router(config-bgp)#address-family ipv4 unicast
RP/0/RSP0/CPU0:router(config-bgp)# neighbor 172.20.2.3 
RP/0/RSP0/CPU0:router(config-bgp-nbr)# remote-as 200
RP/0/RSP0/CPU0:router(config-bgp-nbr)# address-family ipv4 unicast
RP/0/RSP0/CPU0:router(config-bgp-nbr-af)# default-originate

The following example shows how to advertise the route 0.0.0.0/0 to the neighbor 172.20.2.3 only if a route exists in the BGP table that matches the route policy called default-default-policy:

RP/0/RSP0/CPU0:router(config)# router bgp 109
RP/0/RSP0/CPU0:router(config-bgp)# neighbor 172.20.2.3
RP/0/RSP0/CPU0:router(config-bgp-nbr)# remote-as 200
RP/0/RSP0/CPU0:router(config-bgp-nbr)# address-family ipv4 unicast
RP/0/RSP0/CPU0:router(config-bgp-nbr-af)# default-originate route-policy default-default-policy

description (BGP)

To annotate a neighbor, neighbor group, VPN routing and forwarding (VRF) neighbor, or session group, use the description command in an appropriate configuration mode. To remove the annotation, use the no form of this command.

description text

no description [text]

Syntax Description

text

Meaningful description or comment. Maximum of 80 characters.

Command Default

No comment or description exists.

Command Modes

Neighbor group configuration

Neighbor configuration

Session group configuration

VRF neighbor configuration

Command History

Release

Modification

Release 3.7.2

This command was introduced.

Usage Guidelines

To use this command, you must be in a user group associated with a task group that includes appropriate task IDs. If the user group assignment is preventing you from using a command, contact your AAA administrator for assistance.

Use the description command to provide a description of a neighbor, neighbor group, VRF neighbor, or session group. The description is used to save user comments and does not affect software function.

Task ID

Task ID

Operations

bgp

read, write

Examples

The following example shows how to configure the description “Our best customer” on the neighbor 192.168.13.4:


RP/0/RSP0/CPU0:router# configure
RP/0/RSP0/CPU0:router(config)#router bgp 65000 
RP/0/RSP0/CPU0:router(config-bgp)#neighbor 192.168.13.4  
RP/0/RSP0/CPU0:router(config-bgp-nbr)#description Our best customer

distance bgp

To allow the use of external, internal, and local administrative distances that could be used to prefer one class of routes over another, use the distance bgp command in an appropriate configuration mode. To disable the use of administrative distances, use the no no form of this command.

distance bgp external-distance internal-distance local-distance

no distance bgp [external-distance internal-distance local-distance]

Syntax Description

external-distance

Administrative distance for Border Gateway Protocol (BGP) external routes. External routes are routes for which the best path is learned from a neighbor external to the autonomous system. Range is 1 to 255. Routes with a distance of 255 are not installed in the routing table.

internal-distance

Administrative distance for BGP internal routes. Internal routes are those routes that are learned from another BGP entity within the same autonomous system. Range is 1 to 255. Routes with a distance of 255 are not installed in the routing table.

local-distance

Administrative distance for BGP local routes. The local-distance argument applies to locally generated aggregate routes (such as the routes generated by the aggregate-address command) and backdoor routes installed in the routing table. Range is 1 to 255. Routes with a distance of 255 are not installed in the routing table.

Command Default

external-distance : 20

internal-distance : 200

local-distance : 200

Command Modes

IPv4 address family configuration

IPv6 address family configuration

VRF IPv4 address family configuration

VRF IPv6 address family configuration

Command History

Release

Modification

Release 3.7.2

This command was introduced.

Usage Guidelines

To use this command, you must be in a user group associated with a task group that includes appropriate task IDs. If the user group assignment is preventing you from using a command, contact your AAA administrator for assistance.

Use the distance bgp command if another protocol is known to be able to provide a better route to a node than was actually learned using external BGP, or if some internal routes should be preferred by BGP.


Note


Changing the administrative distance of BGP internal routes is considered risky and is not recommended. One problem that can arise is the accumulation of routing table inconsistencies, which can interfere with routing.


An administrative distance is a rating of the trustworthiness of a routing information source. Numerically, an administrative distance is an integer from 1 to 255. In general, the higher the value, the lower the trust rating. An administrative distance of 255 means the routing information source cannot be trusted at all and should be ignored.

Task ID

Task ID

Operations

bgp

read, write

Examples

The following example shows that iBGP routes are preferable to locally generated routes, so the administrative distance values are set accordingly:


RP/0/RSP0/CPU0:router(config)# router bgp 109
RP/0/RSP0/CPU0:router(config-bgp)# address-family ipv4 unicast
RP/0/RSP0/CPU0:router(config-bgp-af)#distance bgp 20 20 200 

distribute bgp-ls (ISIS)

To distribute ISIS link-state data using BGP LS, use the distribute bgp-ls command in router configuration mode. To stop link-state distribution, use the no form of this command.

distribute bgp-ls [instance-id value] [level {1 | 2}] [throttle time]

no distribute bgp-ls

Syntax Description

instance-id value

(Optional) Specifies the instance identifier defined by the router isis command. Range is from 1 to 65535.

If the instance-id is not configured, the system assigned instance-id for the ISIS process will be used.

level 1 | 2

(Optional) Displays IS-IS link-state database for Level 1 or Level 2 independently.

throttle

(Optional) Specifies throttle update, in seconds. Range is from 5 to 20 seconds.

Command Default

None

Command Modes

Router configuration.

Command History

Release

Modification

Release 5.1.1

This command was introduced.

Usage Guidelines

To use this command, you must be in a user group associated with a task group that includes appropriate task IDs. If the user group assignment is preventing you from using a command, contact your AAA administrator for assistance.

Task ID

Task ID

Operations

isis

read, write

Examples

This example shows how to distribute ISIS link-state information using BGP LS:


RP/0/RSP0/CPU0:router# configure
RP/0/RSP0/CPU0:router(config)# router isis foo
RP/0/RSP0/CPU0:router(config-isis)# distribute bgp-ls instance-id 32 level 2 throttle 5

distribute bgp-ls (OSPF)

To distribute OSPFv2 and OSPFv3 link-state data using BGP LS, use the distribute bgp-ls command in router configuration mode. To stop link-state distribution, use the no form of this command.

distribute bgp-ls [instance-id value] [throttle time]

no distribute bgp-ls

Syntax Description

instance-id value

(Optional) Specifies the instance identifier defined by the router ospf command. Range is from 1 to 65535.

If the instance-id is not configured, the system assigned instance-id for the OSPF process is used.

throttle

(Optional) Specifies throttle time between successive link-state advertisement (LSA) updates. Range is from 0 to 3600.

Command Default

BGP distribution is disabled.

Command Modes

Router configuration.

Command History

Release

Modification

Release 5.1.1

This command was introduced.

Usage Guidelines

To use this command, you must be in a user group associated with a task group that includes appropriate task IDs. If the user group assignment is preventing you from using a command, contact your AAA administrator for assistance.

Task ID

Task ID

Operations

ospf

read, write

Examples

This example shows how to distribute OSPF link-state information using BGP LS:


RP/0/RSP0/CPU0:router# configure
RP/0/RSP0/CPU0:router(config)# router ospf 100
RP/0/RSP0/CPU0:router(config-ospf)# distribute bgp-ls instance-id 32 throttle 10

distributed speaker

To enable a distributed speaker process, use the distributed speaker command in router configuration mode. To remove the distributed speaker process, use the no form of this command.

distributed speaker id

no distributed speaker id

Syntax Description

id

ID of the distributed speaker process. Range is 1 to 15.

Command Default

Default is 0.

Command Modes

Router configuration

Command History

Release

Modification

Release 3.7.2

This command was introduced.

Usage Guidelines

To use this command, you must be in a user group associated with a task group that includes appropriate task IDs. If the user group assignment is preventing you from using a command, contact your AAA administrator for assistance.

If BGP is currently running in standalone mode, you must enter the clear bgp current-mode command to switch from standalone or distributed mode.

Task ID

Task ID

Operations

bgp

read, write

Examples

The following example shows how to enable distributed speaker process 3:


RP/0/RSP0/CPU0:router(config)# router bgp 109
RP/0/RSP0/CPU0:router(config-bgp)# distributed speaker 3

domain-distinguisher

To configure globally unique identifier ASN for IGP domain, use the domain-distinguisher command in address-family link-state configuration mode. To remove unique identifier, use the no form of this command.

domain-distinguisher unique-id

no domain-distinguisher

Syntax Description

unique-id

Specifies four-octet unique identifier ASN. Range is from 1 to 4294967295.

Command Default

None

Command Modes

Address-family link-state configuration.

Command History

Release

Modification

Release 5.1.1

This command was introduced.

Usage Guidelines

To use this command, you must be in a user group associated with a task group that includes appropriate task IDs. If the user group assignment is preventing you from using a command, contact your AAA administrator for assistance.

Task ID

Task ID

Operations

bgp

read, write

Examples

This example shows how to configure a unique identifier ASN:


RP/0/RSP0/CPU0:router# configure
RP/0/RSP0/CPU0:router(config)# router bgp 100
RP/0/RSP0/CPU0:router(config-bgp)# address-family link-state link-state
RP/0/RSP0/CPU0:router(config-bgp-af)# domain-distinguisher 1234

dmz-link-bandwidth

To originate a demilitarized zone (DMZ) link bandwidth extended community for the link to an eBGP or iBGP neighbor, use the dmz-link-bandwidth command in an Neighbor configuration mode. To stop origination of the DMZ link bandwidth extended community, use the no form of this command.

dmz-link-bandwidth [ inheritance-disable ]

no dmz-link-bandwidth

Syntax Description

inheritance-disable

(Optional) Prevents the dmz-link-bandwidth command from being inherited from a parent group.

Command Default

BGP does not originate the DMZ link bandwidth extended community.

Command Modes

Neighbor configuration

Command History

Release Modification

Usage Guidelines

To use this command, you must be in a user group associated with a task group that includes appropriate task IDs. If the user group assignment is preventing you from using a command, contact your AAA administrator for assistance.

Use the dmz-link-bandwidth command to advertise the bandwidth of links that are used to exit an autonomous system.

Task ID

Task ID

Operations

bgp

read, write

Examples

This example shows how to advertise the bandwidth of links to eBGP neighbors from router bgp 1:


Router(config)# router bgp 1
Router(config-bgp)#neighbor 45.67.89.01 
Router(config-bgp-nbr)#dmz-link-bandwidth 

dscp (BGP)

To set the differentiated services code point (DSCP) value, use the dscp command in the appropriate configuration mode. To remove the dscp command from the configuration file and restore the system to its default interval values, use the no form of this command.

dscp value

no dscp [value]

Syntax Description

value

Value of the DSCP. The DSCP value can be a number from 0 to 63, or it can be one of the following keywords: default , ef , af11 , af12 , af13 , af21 , af22 , af23 , af31 , af32 , af33 , af41 , af42 , af43 , cs1 , cs2 , cs3 , cs4 , cs5 , cs6 , or cs7 .

Command Default

No default behavior or values

Command Modes

Neighbor configuration

Neighbor session group configuration

Neighbor group configuration

Command History

Release

Modification

Release 3.7.2

This command was introduced.

Usage Guidelines

To use this command, you must be in a user group associated with a task group that includes appropriate task IDs. If the user group assignment is preventing you from using a command, contact your AAA administrator for assistance.

Use the dscp command to change the minimum and maximum packet thresholds for the DSCP value.

Table 1 lists the DSCP default settings used by the dscp command. The DSCP value, corresponding minimum threshold, maximum threshold, and mark probability are listed. The last row of the table (the row labeled "default") shows the default settings used for any DSCP value not specifically shown in the table.

Table 2. dscp Default Settings

DSCP (Precedence)

Minimum Threshold

Maximum Threshold

Mark Probability

af11

32

40

1/10

af12

28

40

1/10

af13

24

40

1/10

af21

32

40

1/10

af22

28

40

1/10

af23

24

40

1/10

af31

32

40

1/10

af32

28

40

1/10

af33

24

40

1/10

af41

32

40

1/10

af42

28

40

1/10

af43

24

40

1/10

cs1

22

40

1/10

cs1

24

40

1/10

cs3

26

40

1/10

cs4

28

40

1/10

cs5

30

40

1/10

cs6

32

40

1/10

cs7

34

40

1/10

ef

36

40

1/10

default

20

40

1/10

Task ID

Task ID

Operations

bgp

read, write

Examples

The following example shows how to set the DSCP value to af32:


RP/0/RSP0/CPU0:router(config)# router bgp 5
RP/0/RSP0/CPU0:router(config-bgp)#neighbor 10.1.1.1  
RP/0/RSP0/CPU0:router(config-bgp-nbr)#remote-as 100 
RP/0/RSP0/CPU0:router(config-bgp-nbr)# dscp af32 

dynamic-med interval

To customize advertisement of Route Distinguisher (RD) updates to neighbors, use the dynamic-med-interval command in BGP address family configuration mode for IPv4, IPv6, VPNv4, and VPNv6 unicast address families. To undo this command configuration, use the no form of this command.

dynamic-med interval interval-value

no dynamic-med interval interval-value

Syntax Description

interval interval-value

Specifies interval (in seconds) between two consecutive update generation to neighbors.

Command Default

The default value of dynamic-med interval is 600 seconds.

Command Modes

BGP address family configuration mode

Command History

Release Modification

Release 5.3.2

This command was introduced.

Task ID

Task ID Operation

bgp

read, write

Examples

The following example shows how to customize rd-advertisement of updates to neighbors:


Router# configure
Router(config)# router bgp 100
Router(config-bgp)# address-family ipv4 unicast
Router(config-bgp-af)# dynamic-med interval 2
Router(config-bgp-nbr-af)# commit

ebgp-multihop

To accept and attempt Border Gateway Protocol (BGP) connections to external peers residing on networks that are not directly connected, use the ebgp-multihop command in an appropriate configuration mode. To disable connections to external peers and allow only direct connections between neighbors, use the no form of this command.

ebgp-multihop [ttl-value] [mpls]

no ebgp-multihop [ttl-value] [mpls]

Syntax Description

ttl-value

(Optional) Time-to-live (TTL) value. Range is 1 to 255 hops.

mpls

(Optional) Disables BGP label rewrite.

Command Default

Default TTL value is 255.

Command Modes

Neighbor configuration

VRF neighbor configuration

Neighbor group configuration

Session group configuration

Command History

Release

Modification

Release 3.7.2

This command was introduced.

Release 4.0.0

The mpls keyword was supported.

Usage Guidelines

To use this command, you must be in a user group associated with a task group that includes appropriate task IDs. If the user group assignment is preventing you from using a command, contact your AAA administrator for assistance.

Use the ebgp-multihop command to enable multihop peerings with external BGP neighbors. The BGP protocol states that external neighbors must be directly connected (one hop away). The software enforces this by default; however, the ebgp-multihop command can be used to override this behavior.

Use of the mpls option in the ebgp-multihop command prevents BGP from enabling MPLS on the peering interface and also prevents allocation of Implicit-NULL rewrite labels for nexthop addresses learned from the peer. This is useful in some scenarios in which MPLS forwarding labels to the nexthops have already been learned via BGP labeled-unicast or LDP.

If this command is configured for a neighbor group or session group, all neighbors using the group inherit the configuration. Values of commands configured specifically for a neighbor override inherited values.

Task ID

Task ID

Operations

bgp

read, write

Examples

The following example shows how to allow a BGP connection to neighbor 172.20.16.6 of up to 255 hops away:


RP/0/RSP0/CPU0:router(config)# router bgp 109
RP/0/RSP0/CPU0:router(config-bgp)# neighbor 172.20.16.6
RP/0/RSP0/CPU0:router(config-bgp-nbr)# ebgp-multihop

export route-policy

To configure an export route policy, use the export route-policy command in an appropriate configuration mode. To restore the system to its default condition, use the no form of this command.

export route-policy policy-name

no export route-policy [policy-name]

Syntax Description

policy-name

Name of the configured route policy.

Command Default

No default behavior or values

Command Modes

Global VRF IPv4 address family configuration

Global VRF IPv6 address family configuration

Command History

Release

Modification

Release 3.7.2

This command was introduced.

Usage Guidelines

To use this command, you must be in a user group associated with a task group that includes appropriate task IDs. If the user group assignment is preventing you from using a command, contact your AAA administrator for assistance.

Use the export route-policy command to define the conditions that allow specified routes to be tagged with specified route-targets.

Task ID

Task ID

Operations

bgp

read, write

ip-services

read, write

Examples

The following example shows how to configure an export route policy:


RP/0/RSP0/CPU0:router(config)# vrf vrf-1
RP/0/RSP0/CPU0:router(config-vrf)#address-family ipv4 unicast  
RP/0/RSP0/CPU0:router(config-vrf-af)# export route-policy policy-A

export route-target

To configure a VPN routing and forwarding (VRF) export route-target extended community, use the export route-target command in an appropriate configuration mode. To restore the system to its default condition, use the no form of this command.

export route-target [as-number:nn | | ip-address:nn]

no export route-target [as-number:nn | | ip-address:nn]

Syntax Description

as-number:nn

(Optional) as-number —Autonomous system (AS) number of the route-target extended community.

  • as-number
    • Range for 2-byte Autonomous system numbers (ASNs) is 1 to 65535.

    • Range for 4-byte Autonomous system numbers (ASNs) in asplain format is 1 to 4294967295.

    • Range for 4-byte Autonomous system numbers (ASNs) is asdot format is 1.0 to 65535.65535.

  • nn —32-bit number

ip-address:nn

(Optional) IP address of the route-target extended community.

  • ip-address —32-bit IP address
  • nn —16-bit number

Command Default

No default behavior or values

Command Modes

Global VRF IPv4 address family configuration

Global VRF IPv6 address family configuration

Command History

Release

Modification

Release 3.7.2

This command was introduced.

Release 3.9.0

Asplain format for 4-byte Autonomous system numbers notation was supported.

Usage Guidelines

To use this command, you must be in a user group associated with a task group that includes appropriate task IDs. If the user group assignment is preventing you from using a command, contact your AAA administrator for assistance.

Export route-target extended communities are associated with prefixes when advertised to remote provider edge (PE) routers. The remote PE routers import the route-target extended communities into a VRF instance that has the import route-targets that match the exported route-target extended communities.

To specify multiple route targets, enter export route target configuration mode then enter one route target for each command line.

Task ID

Task ID

Operations

bgp

read, write

ip-services

read, write

Examples

The following example shows how to specify an export route-target:


RP/0/RSP0/CPU0:router(config)# vrf vrf-1
RP/0/RSP0/CPU0:router(config-vrf)# address-family ipv4 unicast 
RP/0/RSP0/CPU0:router(config-vrf-af)# export route-target 500:1 

graceful-maintenance

To allow the network to perform convergence before the router or link is taken out of service, use the graceful-maintenance command in the router BGP, neighbor or neighbor group configuration mode, as appropriate. To disable the command, use the no form of this command.

graceful-maintenance activate [all-neighbors | retain-routes]


Note


This command is executed in the router BGP configuration mode.


graceful-maintenance {activate [as-prepends as-prepends-value] [inheritance-disable] | [local-preference local-pref-value] inheritance-disable}


Note


This command is executed in either the neighbor configuration or neighbor group configuration mode.

Syntax Description

activate

Announces routes with the graceful maintenance attributes while activated either under the neighbor or router BGP configuration. While activated, all routes to this neighbor are announced with the attribute configured here and all routes from this neighbor are announced to other neighbors with the graceful maintenance attributes configured under those neighbors. The GSHUT community is announced regardless of the other attributes configured here. To allow the GSHUT community to be announced to eBGP neighbors, you must configure the send-community-gshut-ebgp command.

all-neighbors

If you use the all-neighbors keyword, Graceful Maintenance is activated even for those neighbors that do not have Graceful Maintenance activated.

retain-routes

Choosing retain-routes causes RIB to retain BGP routes when the BGP process is stopped. You would use retain-routes when only BGP is being brought down instead of the entire router and if it is known that neighboring routers are being kept in operation during the maintenance of the local BGP. If RIB has alternative routes provided by another protocol or a default route, then it is recommended not to retain BGP routes after the BGP process stops.

as-prepends as-prepends-value inheritance-disable

Indicates the number of times to prepend the local AS number to the AS path of routes. The default value is 0. The keyword inheritance-disable prevents AS prepends from being inherited from the parent.

Specifies the number of times to prepend the local AS number to the AS path of routes and advertises the GSHUT community with the local preference value specified for the routes. When the router adds the GSHUT community to a route as it advertises it, it also changes the LOCAL_PREF attribute and prepends the local AS number as specified in the commands. Sending GSHUT provides flexibility in how neighboring routers handle the lower preference: they can match it in a route policy and do the most appropriate thing with it. On the other hand, in simple networks, it is recommended to set local-preference to 0, rather than to create route policies everywhere else.

Note

 

LOCAL_PREF is not sent to real eBGP neighbors, but sent to confederation member AS eBGP neighbors. To lower preference to eBGP neighbors, as-prepends is required.

local-preference local-pref-value inheritance-disable

Indicates the range of values for Local Preference. The keyword inheritance-disable prevents local preference from being inherited from the parent.

Command Default

None

Command Modes

router BGP

neighbor configuration

neighbor group configuration

Command History

Release

Modification

Release 5.3.2

This command was introduced.

Task ID

Task ID

Operations

bgp

read, write

host-reachability protocol bgp

To configure a VxLAN Tunnel EndPoint (VTEP) with BGP as the control plane that provides BGP protocol-based VTEP peer discovery and end-host reachability information distribution, use the host-reachability protocol bgp command in the VNI NVE configuration mode. To revert to the default configuration, use the no form of this command.

host-reachability protocol bgp

no host-reachability protocol bgp

Syntax Description

This command has no keywords or arguments.

Command Default

The VTEP relies on a multicast-based data plane flood-and-learn behavior for VTEP peer discovery and end-host reachability information as defined in the initial IETF VXLAN standards (RFC 7348).

Command Modes

VNI NVE configuration mode

Command History

Release Modification

Release 5.3.2

This command was introduced.

Usage Guidelines

To use this command, you must be in a user group associated with a task group that includes appropriate task IDs. If the user group assignment is preventing you from using a command, contact your AAA administrator for assistance.

Task ID

Task ID Operation

tunnel, interface

read, write

Examples

The following example configuration shows BGP configured as the control plane for a Network Virtualization EndPoint (NVE) interface (VTEP) that is associated with a VxLAN with identifier 1.


RP/0/RSP0/CPU0:router# configure
RP/0/RSP0/CPU0:router(config)# interface nve 1
RP/0/RSP0/CPU0:router(config-nve)# member vni 1
RP/0/RSP0/CPU0:router(config-nve-vni)# host-reachability protocol bgp 

ibgp policy out enforce-modifications

To allow an outbound route policy for an internal BGP (iBGP) peer to modify all BGP route attributes, only when an iBGP route is sent to another iBGP peer (only on route-reflectors), use the ibgp policy out enforce-modifications command in router configuration mode. To disable this feature, use the no form of this command.

ibgp policy out enforce-modifications

no ibgp policy out enforce-modifications

Syntax Description

This command has no arguments or keywords.

Command Default

ibgp policy out enforce-modifications is disabled.

Command Modes

Router configuration

Command History

Release

Modification

Release 7.8.1

The command scope was expanded to influence next-hop-self behavior, enforce outbound policy modifications, and optimize label processing in BGP.

Refer to the Usage Guidelines for more information.

Release 3.7.2

This command was introduced.

Usage Guidelines

To use this command, you must be in a user group associated with a task group that includes appropriate task IDs. If the user group assignment is preventing you from using a command, contact your AAA administrator for assistance.

Use the ibgp policy out enforce-modifications command to set and modify BGP route attributes for updates to iBGP peers.

If the ibgp policy out enforce-modifications command is configured under router BGP configuration, then all the changes made by the outbound policy for an iBGP peer will be present in an update message sent to the peer.

For VPN or EVPN with Route Reflectors (RRs) configured with a next-hop-self, no label allocation occurs unless the ibgp policy out enforce-modifications option is configured.

In Option B scenarios with the same Route Distinguisher (RD) and using RRs with next-hop-self, for example, ABR, label allocation does not occur unless the ibgp policy out enforce-modifications option is configured.

RIB (Routing Information Base) processing adheres to the label versioning. Labels are processed before RIB processing and then considered for updates. If an RR is configured with the ibgp policy out enforce-modifications option, label processing will occur; otherwise, it will be optimized to update the RIB and process updates without label processing.

Outbound policy modifications on RRs, such as modifying extcomm and largecomm, are only applied when the ibgp policy out enforce-modifications option is configured.

Modifying the AS Path using Prepend and Replace-AS in RRs is effective only with the ibgp policy out enforce-modifications configuration.

For RRs with multipath or next-hop-self, the selected label during sending updates is the local label only when configured with the ibgp policy out enforce-modifications configuration.

Next-hop-self in RRs is applied with the ibgp policy out enforce-modifications configuration.

For 6PE detection where a labeled path is from a V4 Neighbor, implicit next-hop is applied when advertising an IPv6 Unicast. Cross AFI (Address Family Identifier) labeled unicast is not advertised to IPv6 unicast if the ibgp policy out enforce-modifications option is not configured.

Task ID

Task ID

Operations

bgp

read, write

Examples

The following example shows how to set the ibgp policy out enforce-modifications:


RP/0/RSP0/CPU0:routerconfigure
RP/0/RSP0/CPU0:router(config)# router bgp 6500
RP/0/RSP0/CPU0:router(config-bgp)# ibgp policy out enforce-modifications

import

To configure the import and export of BGP NLRIs between two BGP neighbors with respect to the route target identifiers, use the import command in the BGP neighbor address family configuration mode. To undo this command configuration, run the no form of this command.

import [ stitching-rt ] reoriginate [ stitching-rt ]

no import [ stitching-rt ] reoriginate [ stitching-rt ]

Syntax Description

reoriginate

Imports NLRIs that match normal route target identifier and exports re-originated NLRIs assigned with the normal route target identifier.

reoriginate stitching-rt

Imports NLRIs that match normal route target identifier and exports re-originated NLRIs assigned with the stitching route target identifier

stitching-rt reoriginate

Imports NLRIs that match stitching route target identifier and exports re-originated NLRIs assigned with the normal route target identifier.

stitching-rt reoriginate stitching-rt

Imports NLRIs that match stitching route target identifier and exports re-originated NLRIs assigned with the stitching route target identifier.

Command Default

None

Command Modes

BGP neighbour address family configuration mode

Command History

Release Modification

Release 5.3.2

This command was introduced.

Usage Guidelines

To use this command, you must be in a user group associated with a task group that includes appropriate task IDs. If the user group assignment is preventing you from using a command, contact your AAA administrator for assistance.

Task ID

Task ID Operation

bgp

read, write

Examples

The following example shows how to configure DCI router to import BGP NLRIs that match normal route target identifier and to export re-originated BGP NLRIs assigned with the stitching route target identifier.


RP/0/RSP0/CPU0:router# configure
RP/0/RSP0/CPU0:router(config)# router bgp 1
RP/0/RSP0/CPU0:router(config-bgp)# neighbor 1.1.1.1
RP/0/RSP0/CPU0:router(config-bgp-nbr)# address-family l2vpn evpn 
RP/0/RSP0/CPU0:router(config-bgp-nbr-af)# import reoriginate stitching-rt

import route-policy

To configure an import route policy, use the import route-policy command in an appropriate configuration mode. To restore the system to its default condition, use the no form of this command.

import route-policy policy-name

no import route-policy [policy-name]

Syntax Description

policy-name

Name of the configured route policy.

Command Default

No default behavior or values

Command Modes

Global VRF IPv4 address family configuration

Global VRF IPv6 address family configuration

Command History

Release

Modification

Release 3.7.2

This command was introduced.

Usage Guidelines

To use this command, you must be in a user group associated with a task group that includes appropriate task IDs. If the user group assignment is preventing you from using a command, contact your AAA administrator for assistance.

Use the import route-policy command to define the conditions that allow specified routes to be imported into the VPN routing and forwarding (VRF) instance if the routes are tagged with specified route-targets.

Task ID

Task ID

Operations

bgp

read, write

ip-services

read, write

Examples

The following example shows how to allow only policy-B to be imported to VRF:

RP/0/RSP0/CPU0:router(config)# vrf vrf-1
RP/0/RSP0/CPU0:router(config-vrf)# address-family ipv4 unicast 
RP/0/RSP0/CPU0:router(config-vrf-af)# import route-policy policy-B

import route-target

To configure a VPN routing and forwarding (VRF) import route-target extended community, use the import route-target command in an appropriate configuration mode. To restore the system to its default condition, use the no form of this command.

import route-target [as-number:nn | | ip-address:nn]

no import route-target [as-number:nn | | ip-address:nn]

Syntax Description

as-number:nn

(Optional) Autonomous system (AS) number of the route-target extended community.

  • as-number
    • Range for 2-byte Autonomous system numbers (ASNs) is 1 to 65535.

    • Range for 4-byte Autonomous system numbers (ASNs) in asplain format is 1 to 4294967295.

    • Range for 4-byte Autonomous system numbers (ASNs) is asdot format is 1.0 to 65535.65535.

  • nn —32-bit number

ip-address:nn

(Optional) IP address of the route-target extended community.

  • ip-address —32-bit IP address
  • nn —16-bit number

Command Default

No default behavior or values

Command Modes

Global VRF IPv4 address family configuration

Global VRF IPv6 address family configuration

Command History

Release

Modification

Release 3.7.2

This command was introduced.

Release 3.9.0

Asplain format for 4-byte Autonomous system numbers notation was supported.

Usage Guidelines

To use this command, you must be in a user group associated with a task group that includes appropriate task IDs. If the user group assignment is preventing you from using a command, contact your AAA administrator for assistance.

Use the import route-target command to specify that prefixes associated with the configured import route-target extended communities are imported into the VRF instance.

To specify multiple route targets, enter import route target configuration mode, then enter one route target for each command line.

Task ID

Task ID

Operations

bgp

read, write

ip-services

read, write

Examples

The following example shows how to specify an import route-target:


RP/0/RSP0/CPU0:router(config)#vrf vrf-1 
RP/0/RSP0/CPU0:router(config-vrf)# address-family ipv4 unicast 
RP/0/RSP0/CPU0:router(config-vrf-af)# import route-target 500:99

ignore-connected-check

To enable the software to bypass the directly connected next hop check for single-hop eBGP peering, use the ignore-connected-check command in an appropriate configuration mode. To re-enable the directly connected next hop check, use the no form of this command.

ignore-connected-check [inheritance-disable]

no ignore-connected-check

Syntax Description

inheritance-disable

Prevents the ignore-connected-check command from being inherited from the parent.

Command Default

Ability to bypass the directly connected next hop check is disabled.

Command Modes

Neighbor configuration

Neighbor group configuration

Session group configuration

Command History

Release

Modification

Release 3.9.0

This command was introduced.

Usage Guidelines

To use this command, you must be in a user group associated with a task group that includes appropriate task IDs. If the user group assignment is preventing you from using a command, contact your AAA administrator for assistance.

Task ID

Task ID

Operations

bgp

read, write

Examples

The following example shows how to enable ignore-connected check configuration for neighbor 10.2.3.4:


RP/0/RSP0/CPU0:router(config)# router bgp 100
RP/0/RSP0/CPU0:router(config-bgp)# neighbor 10.2.3.4
RP/0/RSP0/CPU0:router(config-bgp-nbr)# ignore-connected-check

keychain

To apply key chain-based authentication on a TCP connection between two Border Gateway Protocol (BGP) neighbors, use the keychain command in an appropriate configuration mode. To disable key chain authentication, use the no form of this command.

keychain name

no keychain [name]

Syntax Description

name

Key chain name configured using the keychain command. The name must be a maximum of 32 alphanumeric characters.

Command Default

When this command is not specified in the appropriate configuration mode, key chain authentication is not enabled on a TCP connection between two BGP neighbors.

Command Modes

Neighbor configuration

Neighbor group configuration

Session group configuration

Command History

Release

Modification

Release 3.7.2

This command was introduced.

Usage Guidelines

To use this command, you must be in a user group associated with a task group that includes appropriate task IDs. If the user group assignment is preventing you from using a command, contact your AAA administrator for assistance.

Specify a key chain to enable key chain authentication between two BGP peers. Use the keychain command to implement hitless key rollover for authentication.

If this command is configured for a neighbor group or a session group, a neighbor using the group inherits the configuration. Values of commands configured specifically for a neighbor override inherited values.

Note


BGP only supports HMAC-MD5 and HMAC-SHA1-12 cryptographic algorithms.


Task ID

Task ID

Operations

bgp

read, write

Examples

The following example shows how to configure neighbor 172.20.1.1 to use the key chain authentication configured in the keychain_A key chain:

RP/0/RSP0/CPU0:router(config)# router bgp 140 
RP/0/RSP0/CPU0:router(config-bgp)# neighbor 172.20.1.1 
RP/0/RSP0/CPU0:router(config-bgp-nbr)# remote-as 1 
RP/0/RSP0/CPU0:router(config-bgp-nbr)# keychain keychain_A  

keychain-disable


Note


Effective with Release 3.9.0, the keychain-disable command was replaced by the keychain inheritance-disable command. See the keychain inheritance-disable command for more information.

To override any inherited key chain configuration from a neighbor group or session group for Border Gateway Protocol (BGP) neighbors, use the keychain-disable command in an appropriate configuration mode. To disable overriding any inherited key chain command, use the no form of this command.

keychain-disable

no keychain-disable

Syntax Description

This command has no arguments or keywords.

Command Default

Configured key chains for neighbor and session groups are inherited.

Command Modes

Neighbor configuration

Neighbor group configuration

Session group configuration

Command History

Release

Modification

Release 3.7.2

This command was introduced.

Release 3.9.0

This command was replaced by the keychain inheritance-disable command.

Usage Guidelines

To use this command, you must be in a user group associated with a task group that includes appropriate task IDs. If the user group assignment is preventing you from using a command, contact your AAA administrator for assistance.

If you specify a key chain on a neighbor group or session group, all users of the group inherit the key chain. Specifying a different keychain command specifically on a neighbor that uses the group overrides the inherited value. Specifying keychain-disable on a neighbor that uses the group disables key chain authentication for the neighbor.

Task ID

Task ID

Operations

bgp

read, write

Examples

The following example shows how to disable key chain authentication for neighbor 172.20.1.1, preventing it from inheriting the key chain keychain_A from session group group1:


RP/0/RSP0/CPU0:router(config)# router bgp 140 
RP/0/RSP0/CPU0:router(config-bgp)# session-group group1 
RP/0/RSP0/CPU0:router(config-bgp-sngrp)# keychain keychain_A 
RP/0/RSP0/CPU0:router(config-bgp-sngrp)# exit 
RP/0/RSP0/CPU0:router(config-bgp)#neighbor 172.20.1.1  
RP/0/RSP0/CPU0:router(config-bgp-nbr)# remote-as 2 
RP/0/RSP0/CPU0:router(config-bgp-nbr)#use session-group group1  
RP/0/RSP0/CPU0:router(config-bgp-nbr)# keychain-disable 

keychain inheritance-disable

To override any inherited key chain configuration from a neighbor group or session group for Border Gateway Protocol (BGP) neighbors, use the keychain inheritance-disable command in an appropriate configuration mode. To disable overriding any inherited key chain command, use the no form of this command.

keychain inheritance-disable

no keychain inheritance-disable

Syntax Description

This command has no arguments or keywords.

Command Default

Configured key chains for neighbor and session groups are inherited.

Command Modes

Neighbor configuration

Neighbor group configuration

Session group configuration

Command History

Release

Modification

Release 3.9.0

This command was introduced.

Usage Guidelines

To use this command, you must be in a user group associated with a task group that includes appropriate task IDs. If the user group assignment is preventing you from using a command, contact your AAA administrator for assistance.

If you specify a key chain on a neighbor group or session group, all users of the group inherit the key chain. Specifying a different keychain command specifically on a neighbor that uses the group overrides the inherited value. Specifying keychain inheritance-disable on a neighbor that uses the group disables key chain authentication for the neighbor.

Task ID

Task ID

Operations

bgp

read, write

Examples

The following example shows how to disable key chain authentication for neighbor 172.20.1.1, preventing it from inheriting the key chain keychain_A from session group group1:

RP/0/RSP0/CPU0:router(config)#router bgp 140 
RP/0/RSP0/CPU0:router(config-bgp)# session-group group1 
RP/0/RSP0/CPU0:router(config-bgp-sngrp)# keychain keychain_A 
RP/0/RSP0/CPU0:router(config-bgp-sngrp)# exit 
RP/0/RSP0/CPU0:router(config-bgp)# neighbor 172.20.1.1 
RP/0/RSP0/CPU0:router(config-bgp-nbr)# remote-as 2 
RP/0/RSP0/CPU0:router(config-bgp-nbr)# use session-group group1 
RP/0/RSP0/CPU0:router(config-bgp-nbr)# keychain inheritance-disable 

label-allocation-mode

To set the MPLS/VPN label allocation mode, use the label-allocation-mode command in VRF configuration mode. To remove the label-allocation-mode command from the configuration file and restore the system to its default condition, use the no form of this command.

label-allocation-mode [per-ce | ]

no label-allocation-mode

Syntax Description

per-ce

Specifies that the same label is used for all the routes advertised from a unique customer edge (CE) peer or router.

Command Default

Per-prefix is the default label allocation mode.

Command Modes

VRF configuration

Command History

Release

Modification

Release 3.7.2

This command was introduced.

Release 7.4.1

This command was deprecated


Note


The function of this deprecated command can be carried out using label mode command under configured address-family.


Usage Guidelines

To use this command, you must be in a user group associated with a task group that includes appropriate task IDs. If the user group assignment is preventing you from using a command, contact your AAA administrator for assistance.

Each prefix that belongs to a VRF instance is advertised with a single label, causing an additional lookup to be performed in the VRF forwarding table to determine the customer edge (CE) next hop for the packet. Use the label-allocation-mode command with the per-ce keyword to avoid the additional lookup on the PE router and conserve label space. This mode allows the PE router to allocate one label for every immediate next hop. The label is directly mapped to the next hop so there is no VRF route lookup performed during data forwarding. However, the number of labels allocated is one for each CE rather than one for each prefix.

Task ID

Task ID

Operations

bgp

read, write

Examples

The following example shows how to set the label allocation mode to customer edge:

RP/0/RSP0/CPU0:router(config)# router bgp 109
RP/0/RSP0/CPU0:router(config-bgp)# vrf vrf-1
RP/0/RSP0/CPU0:router(config-bgp-vrf)# label-allocation-mode per-ce 

label mode

To set the MPLS/VPN label mode based on prefix value, use the label mode command in an appropriate configuration mode. To remove the label mode command from the configuration file and restore the system to its default condition, use the no form of this command.

Use this syntax for vrf all configuration mode under VPN IPv4/IPv6 AF (address family) mode or global IPv6 AF configuration mode:

label mode { per-ce | | per-vrf | | route-policy }

no label mode { per-ce | | per-vrf | | route-policy }

Use this syntax for IPv4/IPv6 AF configuration mode under vrf mode:

label mode { per-prefix | | per-ce | | per-vrf | | route-policy }

no label mode { per-prefix | | per-ce | | per-vrf | | route-policy }

Syntax Description

per-ce

Specifies that the same label is used for all routes advertised from a unique customer edge (CE) peer or route.

per-vrf

Specifies that the same label is used for all routes advertised from a unique VRF.

per-prefix

Specifies that the same label is used for all routes advertised from a unique prefix.

Note

 

This keyword is applicable only for IPv4/IPv6 AF configuration mode under vrf mode.

route-policy

Specifies a route policy to select prefixes for setting the label mode.

Command Default

Per-prefix label mode.


Note


If a policy attached at label-mode attachpoint evaluates to pass and a label mode is not explicitly set, per-prefix is used as the default label mode.

If a policy attached at label-mode attachpoint evaluates to a drop, per-prefix is used as a default label mode. If any label mode is set explicitly in this case, it will be ignored.


Command Modes

VPNv4 address family configuration

VPNv6 address family configuration

VRF IPv4 address family configuration

VRF IPv6 address family configuration

Command History

Release Modification

Release 4.3.1

This command was introduced.

Usage Guidelines

To use this command, you must be in a user group associated with a task group that includes appropriate task IDs. If the user group assignment is preventing you from using a command, contact your AAA administrator for assistance.

To configure label mode at VPN-AF level and to have all the VRF AFs inherit that configuration, you must use vrf all , which is available under VPN-AF mode.

The inheritance rules followed are:

  • label mode configuration under VRF-AF, overrides label-allocation-mode configuration under VRF and label mode configuration under VPN-AF.

  • label-allocation-mode configuration under VRF, overrides label mode configuration under VPN-AF.

  • The order of priority to determine the label mode in the configurations is:

    1. VRF-AF: label mode

    2. VRF: label-allocation-mode

    3. VPN-AF: label mode

    4. N/A: per-prefix


Note


Even if label mode is in use, per-vrf label is allocated for connected, aggregate, and local prefixes.


Task ID

Task ID Operation

bgp

read, write

Examples

The example shows how to configure label mode selection at VPNv4 AF level:


Router# configure
Router(config)# router bgp 100
Router(config-bgp)# address-family vpnv4 unicast
Router(config-bgp-af)# vrf all
Router(config-bgp-af)# label mode route-policy policy_A

The example shows how to configure label mode selection at VRF IPv4 AF level:


Router# configure
Router(config)# router bgp 109
Router(config-bgp)# vrf vrf-1
Router(config-bgp-vrf)# rd 1:1
Router(config-bgp-vrf)# address-family ipv4 unicast
Router(config-bgp-vrf)# label mode route-policy policy_B

local-as

To allow customization of the autonomous system number for external Border Gateway Protocol (eBGP) neighbor peerings, use the local-as command in an appropriate configuration mode. To disable customization of local autonomous system values for eBGP neighbor peerings, use the no form of this command.

local-as {as-number [no-prepend [replace-as [dual-as]]] | inheritance-disable}

no local-as [as-number [no-prepend [replace-as [dual-as]]] | inheritance-disable]

Syntax Description

as-number

Valid autonomous system number.

Range for 2-byte Autonomous system numbers (ASNs) is 1 to 65535.

Range for 4-byte Autonomous system numbers (ASNs) in asplain format is 1 to 4294967295.

Range for 4-byte Autonomous system numbers (ASNs) is asdot format is 1.0 to 65535.65535.

Cannot be the autonomous system number to which the neighbor belongs.

no-prepend

(Optional) Specifies that local autonomous system values are not prepended to announcements from the neighbor.

replace-as

(Optional) Specifies that prepend only local autonomous system values to announcements to the neighbor.

dual-as

(Optional) Dual-AS mode.

inheritance-disable

Prevents local AS from being inherited from the parent.

Command Default

The BGP autonomous system number specified in the router bgp command is used, except when confederations are in use. The confederation autonomous system is used for external neighbors in an autonomous system that is not part of the confederation.

Command Modes

Neighbor configuration

VRF neighbor configuration

Neighbor group configuration

Session group configuration

Command History

Release

Modification

Release 3.7.2

This command was introduced.

Release 3.9.0

Asplain format for 4-byte Autonomous system numbers notation was supported. The dual-as keyword was added. The disable keyword was replaced with the inheritance-disable keyword.

Release 5.2.2

Support was added to specify the same autonomous system number for local-as and remote-as commands.

Usage Guidelines

To use this command, you must be in a user group associated with a task group that includes appropriate task IDs. If the user group assignment is preventing you from using a command, contact your AAA administrator for assistance.

You can specify the autonomous system number the local BGP uses to peer with each neighbor. The autonomous system number specified with this command cannot be the local BGP autonomous system number (specified with the router bgp command) or the autonomous system number of the neighbor (specified with the remote-as command). This command cannot be specified for internal neighbors or for external neighbors in an autonomous system that is part of a confederation.

If this command is configured for a neighbor group or session group, all neighbors using the group inherit the configuration. Values of commands configured specifically for a neighbor override inherited values.

In case the router is an ASBR and adding the local-as configuration changes the BGP neighbor from being an eBGP to iBGP, incremental addition of the local-as command is not supported. Remove the neighbor configuration, then configure the entire neighbor, including remote-as configuration and the new local-as configuration, in one commit.

Task ID

Task ID

Operations

bgp

read, write

Examples

The following example shows BGP using autonomous system 30 for the purpose of peering with neighbor 172.20.1.1:

RP/0/RSP0/CPU0:router(config)# router bgp 140
RP/0/RSP0/CPU0:router(config-bgp)# neighbor 172.20.1.1
RP/0/RSP0/CPU0:router(config-bgp-nbr)# remote-as 300
RP/0/RSP0/CPU0:router(config-bgp-nbr)# local-as 30

long-lived-graceful-restart

To enable long lived graceful restart (LLGR) on the BGP neighbors, use the long-lived-graceful-restart command in neighbor VPN address family mode. To disable LLGR, use the no form of this command.

long-lived-graceful-restart {capable | stale-time send time accept time}

Syntax Description

capable

Treats the neighbor as LLGR capable even if it does not advertise the capabilities.

stale-time

Causes the local router to advertise the LLGR capability to the neighbor and to enable LLGR for prefixes received from the neighbor.

send time

Specifies stale-time sent in LLGR capability.

accept time

Specifies maximum stale-time acceptable from neighbor.

Command Default

The default send and accept time is zero.

Command Modes

VPNv4 address family configuration

VPNv6 address family configuration

Command History

Release

Modification

Release 5.2.2

This command was introduced.

Usage Guidelines

When this command is configured, the BGP session is reset, because the changes need to be advertised to the neighbor in a BGP OPEN message.

When the BGP session to a neighbor goes down the routes received from it will be marked LLGR stale if all of the following conditions are met:
  • Either the neighbor is configured as capable or the neighbor sent the LLGR capability in its BGP OPEN message

  • The accept time is not configured to be 0.

  • The stale time that the neighbor sent in the LLGR capability in its BGP OPEN message is not 0.

  • The neighbor session was not brought down with a clear command on the local router.

  • The neighbor sent either the LLGR or graceful restart capability in its BGP OPEN message.

LLGR routes will only be advertised to a neighbor that is LLGR capable, either because it is configured as capable or because it has sent the LLGR capability in its BGP OPEN message. An LLGR route is either one that has been marked as LLGR stale, because the BGP session from which it was received went down or because it has the LLGR_STALE community and does not have the NO_LLGR community.

Task ID

Task ID

Operations

bgp

read

Examples

This example shows how to configure the neighbor to be LLGR capable for the given address family:


RP/0/RSP0/CPU0:router# configure
RP/0/RSP0/CPU0:router(config)# router bgp 100
RP/0/RSP0/CPU0:router(config-bgp)# neighbor 3.3.3.3
RP/0/RSP0/CPU0:router(config-bgp-nbr)# address-family vpnv4 unicast
RP/0/RSP0/CPU0:router(config-bgp-nbr-af)# long-lived-graceful-restart capable


The long-lived-graceful-restart capable command enables the LLGR capability on the neighbor; even though the neighbor does not advertise the LLGR capabilities during session information.

The following example shows how to advertise :


RP/0/RSP0/CPU0:router# configure
RP/0/RSP0/CPU0:router(config)# router bgp 100
RP/0/RSP0/CPU0:router(config-bgp)# neighbor 3.3.3.3
RP/0/RSP0/CPU0:router(config-bgp-nbr)# address-family vpnv4 unicast
RP/0/RSP0/CPU0:router(config-bgp-nbr-af)# long-lived-graceful-restart stale-time send 20 accept 30

The long-lived-graceful-restart stale-time send 20 accept 30 command is used to configure the LLGR on the neighbor. When this command is configured the configured device will retain routes from the neighbor.

lpts punt excessive-flow-trap routing-protocols-enable

To enable the Excessive Punt Flow Trap (EPFT) on routing protocol packets OSPF and BGP, use the lpts punt excessive-flow-trap routing-protocol-enable command in the Global Configuration mode. To disable, use the no form of this command.

lpts punt excessive-flow-trap routing-protocols-enable

no lpts punt excessive-flow-trap routing-protocols-enable

Command Default

None

Command Modes

Global Configuration mode

Command History

Release Modification

Release 6.0.1

This command was introduced.

Usage Guidelines

You must be in a user group associated with a task group that includes the proper task IDs. The command reference guides include the task IDs required for each command. If you suspect user group assignment is preventing you from using a command, contact your AAA administrator for assistance.

When subscriber interface or interface-based-flow is configured, you can not configure the routing-protocol-enable command. The reverse of this also holds good, that is, if the routing-protocol-enable command is configured, you cannot configure a subscriber interface or interface-based-flow.

L3 routing-protocols to be enabled on non-subscriber interfaces mac.


Note


The routing-protocols-enable command can be configured only after configuring non-subscriber-interface mac.


Task ID

Task ID

Operations

lpts

read

basic-services

read-write

Examples

This example shows to enable EPFT on L3 routing protocols:

RP/0/RSP0/CPU0:router(config)# lpts punt excessive-flow-trap
RP/0/RSP0/CPU0:router(config-control-plane-policer)#  non-subscriber-interfaces mac 
RP/0/RSP0/CPU0:router(config-control-plane-policer)#  routing-protocols-enable 

lpts punt excessive-flow-trap penalty-timeout bgp

To set the penalty timeout for the bgp protocol, use the lpts punt excessive-flow-trap penalty-timeout bgp command in the Global Configuration mode. To restore the default penalty timeout value, use the no form of this command.

lpts punt excessive-flow-trap {penalty-timeout bgp} timeout

no lpts punt excessive-flow-trap {penalty-timeout bgp}

Syntax Description

timeout

The penalty timeout value for the bgp protocol in minutes. It is the period of time at which bgp bad flow remains to be in bad actor state.

Value ranges from 1 to 1000.

Command Default

The default penalty timeout value is 15 minutes.

Command Modes

Global Configuration mode

Command History

Release Modification
Release 6.0.1

This command was introduced.

Usage Guidelines

You must be in a user group associated with a task group that includes the proper task IDs. The command reference guides include the task IDs required for each command. If you suspect user group assignment is preventing you from using a command, contact your AAA administrator for assistance.

If the penalty-timeout value for bgp is configured as 20, then bgp packets are dropped for 20 minutes.

Task ID

Task ID

Operations

lpts

read

basic-services

read-write

Examples

This example shows to set penalty time out for bgp bad actor:

RP/0/RSP0/CPU0:router(config)# lpts punt excessive-flow-trap
RP/0/RSP0/CPU0:router(config-control-plane-policer)#  penalty-timeout bgp <1-1000>

match flow-tag

To identify specific flow-tag values as match criteria in a class-map, use the match flow-tag command in class-map configuration mode. To remove a specific flow tag value from the matching criteria for a class-map, use the no form of this command.

match flow-tag {flow-tag number | }

no match flow-tag {flow-tag number | }

Syntax Description

flow-tag number

A flow-tag number. Range is from 1 to 63.

Command Default

No match criteria is specified.

Command Modes

Class-map configuration

Command History

Release Modification

Release 5.2.2

This command was introduced.

Usage Guidelines

To use this command, you must be in a user group associated with a task group that includes appropriate task IDs. If the user group assignment is preventing you from using a command, contact your AAA administrator for assistance.

Task ID

Task ID Operation
class-map

read, write

Examples

This example uses theshow route command to display the flow-tag for a given route.

RP/0/0/CPU0:ios-xr#
RP/0/0/CPU0:ios-xr#show route 4.4.4.0/24 detail
Thu Nov 14 15:32:21.010 PST
Routing entry for 4.4.4.0/24
Known via "bgp 200", distance 20, metric 0
Tag 300, type external
Installed Nov 14 09:36:55.066 for 05:55:26
Routing Descriptor Blocks
3.3.3.3, from 3.3.3.3, BGP external
Route metric is 0
Label: None
Tunnel ID: None
Extended communities count: 0
NHID:0x0(Ref:0)
Route version is 0x1 (1)
No local label
IP Precedence: Not Set
QoS Group ID: Not Set
Flow-tag: 220
Route Priority: RIB_PRIORITY_RECURSIVE (12) SVD Type
RIB_SVD_TYPE_LOCAL
Download Priority 4, Download Version 7
No advertising protos.
RP/0/0/CPU0:ios-xr#

maximum-paths (BGP)

To control the maximum number of parallel routes that Border Gateway Protocol (BGP) installs in the routing table, use the maximum-paths command in an appropriate configuration mode. To set the maximum number of parallel routes the software installs to the default value, use the no form of this command.

maximum-paths { { ebgp | ibgp } maximum [ unequal-cost ] | eibgp maximum [ equal-cost ] } route-policy route-policy-name [ selective ] }

Syntax Description

ebgp

Specifies external BGP multipath peers.

ibgp

Specifies internal BGP multipath peers.

eibgp

Specifies internal and external BGP multipath peers. eiBGP allows simultaneous use of internal and external paths.

maximum

Maximum number of parallel routes that BGP installs in the routing table. Range is 2 to 8

unequal-cost

(Optional) Allows iBGP multipaths to have different BGP next-hop Interior Gateway Protocol (IGP) metrics. This option is available when either the ibgp or ebgp keywords are used.

equal-cost

(Optional) Allows eiBGP multipaths to have same BGP next-hop Interior Gateway Protocol (IGP) metrics. This option is available when the eibgp keyword is used.

route-policy

(Optional) Specifies the route policy to select multipath.

route-policy-name

(Optional) Name of the route policy.

selective

(Optional) Allow multipaths only from marked neighbors.

Command Default

One path is installed in the routing table.

Command Modes

IPv4 address family configuration

IPv6 address family configuration

VRF IPv4 address family configuration

VRF IPv6 address family configuration

Command History

Release

Modification

Release 3.7.2

This command was introduced.

Release 7.10.1

This command was modified. The equal-cost, route-policy keywords and route-policy-name argument were added.

Usage Guidelines

Use the maximum-paths command to allow the BGP protocol to install multiple paths into the routing table for each prefix. Multiple paths are installed for external peers that are from the same autonomous system and are equal cost (according to the BGP best-path algorithm). Similarly, multiple paths are installed for internal peers that are equal cost based on the BGP best-path algorithm. The IGP metric to the BGP next hop is the same as the best-path IGP metric unless the router is configured for unequal cost iBGP multipath or eiBGP multipath.

Use the equal-cost keyword to maintain consistent nexthop IGP metric criteria for multipath eligibility among iBGP path types.

See Implementing BGP in the BGP Configuration Guide for ASR 9000 Series Routers for information on the BGP best-path algorithm.


Note


The maximum-paths command with the eibgp keyword cannot be configured if the ibgp or ebgp keywords have been configured, because the eibgp keyword is a superset of the ibgp or ebgp keywords.


Task ID

Task ID

Operations

bgp

read, write

Examples

The following example shows how to allow a maximum of four paths to a destination installed into the IPv4 unicast routing table:


Router(config)# router bgp 109
Router(config-bgp)# address-family ipv4 unicast
Router(config-bgp-af)# maximum-paths ebgp 4

Examples

The following example shows to how to preserve consistent nexthop IGP metric criteria for multipath eligibility among similar path types, either iBGP or eBGP.

Router(config)# router bgp 109
Router(config-bgp)# address-family ipv4 unicast
Router(config-bgp-af)# maximum-paths eibgp 32 equal-cost route-policy EIBGP

maximum-prefix (BGP)

To control how many prefixes can be received from a neighbor, use the maximum-prefix command in an appropriate configuration mode. To set the prefix limits to the default values, use the no form of this command.

maximum-prefix maximum [threshold] [warning-only] [restart time-interval]

no maximum-prefix maximum [threshold] [warning-only] [restart time-interval]

Syntax Description

maximum

Maximum number of prefixes allowed from this neighbor. Range is from1 to 4294967295.

threshold

(Optional) Integer specifying at what percentage of the maximum argument value the software starts to generate a warning message. Range is from1 to 100.

warning-only

(Optional) Instructs the software to only generate a log message when the maximum argument value is exceeded, and not to terminate the peering.

restart time-interval

(Optional) Sets the time interval (in minutes) after which peering session should be reestablished.

Configure restart time interval in minutes. Range is from 1 to 65535.

Command Default

When this command is not specified, the following defaults apply:

  • IPv4 Unicast: 1048576

  • IPv4 Labeled-unicast: 131072

  • IPv6 Unicast: 524288

  • IPv6 Labeled-unicast: 131072

  • IPv4 Tunnel: 1048576

  • IPv4 Multicast: 131072

  • IPv6 Multicast: 131072

  • IPv4 MVPN: 2097152

  • VPNv4 Unicast: 2097152

  • IPv4 MDT: 131072

  • VPNv6 Unicast: 1048576

  • L2VPN EVPN: 2097152

  • IPv4 Flowspec: 1048576

  • IPv6 Flowspec: 524288

  • VPNv4 Flowspec: 2097152

  • VPNv6 Flowspec: 1048576

The default threshold, when a warning message is generated, is 75 percent.


Note


Starting from Cisco IOS XR Release 7.3.1, the command has been modified to remove the default maximum-prefix value. You must configure the maximum-prefix value.


Command Modes

IPv4 address family group, neighbor address family, and neighbor group address family configuration

IPv6 address family group, neighbor address family, and neighbor group address family configuration

IPv4 tunnel address family group, neighbor group address family, and neighbor address family configuration

IPv4 flowspec under neighbor address family, neighbor group address family, and address family group configuration

IPv6 flowspec under neighbor address family, neighbor group address family, and address family group configuration

VPNv4 flowspec under neighbor address family, neighbor group address family, and address family group configuration

VPNv6 flowspec under neighbor address family, neighbor group address family, and address family group configuration

L2VPN EVPN under neighbor address family, neighbor group address family, and address family group configuration

Command History

Release

Modification

Release 3.7.2

This command was introduced.

Release 7.3.1

The command has been modified to remove the default maximum-prefix value.

Usage Guidelines

Use the maximum-prefix command to configure a maximum number of prefixes that a BGP router is allowed to receive from a neighbor. It adds another mechanism (besides routing policy) to control prefixes received from a peer.

When the number of received prefixes exceeds the maximum number configured, the software terminates the peering, by default, after sending a cease notification to the neighbor. However, if the warning-only keyword is configured, the software writes only a log message, but continues peering with the sender. If the peer is terminated, the peer stays down until the clear bgp command is issued or the restart time-interval option is used.

This command takes effect immediately if configured on an established neighbor, unless the number of prefixes received from the neighbor already exceeds the configured limits.

If this command is configured for a neighbor group or neighbor address family group, all neighbors using the group inherit the configuration. Values of commands configured specifically for a neighbor override inherited values.

Task ID

Task ID

Operations

bgp

read, write

Examples

This example shows the maximum number of IP Version 4 (IPv4) unicast prefixes allowed from the neighbor 192.168.40.24 set to 1000:

RP/0/RSP0/CPU0:router(config-bgp)# router bgp 109
RP/0/RSP0/CPU0:router(config-bgp)# neighbor 192.168.40.24
RP/0/RSP0/CPU0:router(config-bgp-nbr)# remote-as 1
RP/0/RSP0/CPU0:router(config-bgp-nbr)#address-family ipv4 unicast
RP/0/RSP0/CPU0:router(config-bgp-nbr-af)# maximum-prefix 1000

mpls activate (BGP)

To enable Multiprotocol Label Switching (MPLS) on an interface basis for ASBR and CSC configurations whenever a bgp confederation configuration is used, use the mpls activate command in bgp configuration mode. This is needed for InterAS (option B and C) and Carrier Supporting Carrier (CSC) configurations with confederations.

The normal InterAS and CSC configurations (without confederations) do not need to enable this.

To restore the system to its default condition, use the no form of this command.

mpls activate interface id

no mpls activate interface id

Syntax Description

interface id

Name of the interface.

Command Default

No default behavior or values

Command Modes

Router configuration

Neighbor configuration

IPv4 address family group configuration

VPNv4 address family group configuration

Command History

Release

Modification

Release 3.7.2

This command was introduced.

Usage Guidelines

To use this command, you must be in a user group associated with a task group that includes appropriate task IDs. If the user group assignment is preventing you from using a command, contact your AAA administrator for assistance.

Thempls activate command enables MPLS on the interface specified and also adds the implicit null rewrite corresponding to the peer associated with the interface. The interface specified must be the one corresponding to the inter-AS ASBR or CSC peer.

Task ID

Task ID

Operations

bgp

read, write

Examples

The following example shows how to activate MPLS for InterAS Option B (with confederations):


RP/0/RSP0/CPU0:router(config)#router bgp 1 
  
            
  bgp confederation peers
  
     2002
  
    !
  
    bgp confederation identifier 4589
  
    bgp router-id 3.3.3.3
  
    mpls activate
  
     interface GigabitEthernet0/1/0/0
  
    !
  
    address-family ipv4 unicast
  
     redistribute connected
  
    !
  
    address-family vpnv4 unicast
  
     retain route-target all
  
    !
  
    neighbor 10.0.0.9
  
     remote-as 2002
  
     address-family ipv4 unicast
  
      route-policy pass in
  
      route-policy pass out
  
     !
  
     address-family vpnv4 unicast
  
      route-policy pass in
  
  

The following example shows how to activate MPLS for CSC (with confederations):


  router bgp 2002
    bgp confederation peers
     1
    !
    bgp confederation identifier 4589
    bgp router-id 4.4.4.4
    address-family ipv4 unicast
     allocate-label all
    !
    address-family vpnv4 unicast
     retain route-target all
    !
    vrf foo
     rd 1:1
     mpls activate
      interface GigabitEthernet0/1/0/2
     !
     address-family ipv4 unicast
      redistribute connected
      allocate-label all
     !
     neighbor 10.0.0.1
      remote-as 1
      address-family ipv4 unicast
      !
      address-family ipv4 labeled-unicast
       route-policy pass in
       route-policy pass out
      !
     !
    !
  !
RP/0/RSP0/CPU0:router#show mpls forwarding
  Local  Outgoing    Prefix             Outgoing Next Hop        Bytes
  Label  Label       or ID              Interface 
  Switched
  ------ ----------- ------------------ ------------ ---------------
  ------------
  16000  Aggregate   foo: Per-VRF Aggr[V]   \
                                         foo                          0
  16001  Pop         10.0.0.0/16[V]     Gi0/1/0/2    10.0.0.1        44
  
  
  
  
RP/0/RSP0/CPU0:router#show mpls interfaces
  Interface                  LDP      Tunnel   Enabled
  -------------------------- -------- -------- --------
  GigabitEthernet0/1/0/2     No       No       Yes
  
  

mvpn

To enable BGP instance to connect to PIM/PIM6, use the mvpn command in router configuration mode. To disable BGP instance -PIM/PIM6 connection, use the no form of this command.

mvpn

no mvpn

Syntax Description

This command has no keywords or arguments.

Command Default

PIM/PIM connection is disabled.

Command Modes

Router configuration

Command History

Release Modification

Release 4.2.0

This command was introduced.

Usage Guidelines

To use this command, you must be in a user group associated with a task group that includes appropriate task IDs. If the user group assignment is preventing you from using a command, contact your AAA administrator for assistance.

Task ID

Task ID Operation

bgp

read, write

Examples

This example shows how to configure mvpn and enable PIM/PIM6 connection:


RP/0/RSP0/CPU0:router#configure
RP/0/RSP0/CPU0:router(config)#router bgp 100
RP/0/RSP0/CPU0:router(config-bgp)#mvpn

multipath

Enables multiple paths for a BGP neighbor.

To disable this function, use the no form of this command.

multipath

no multipath

Command Default

Multipath is disabled by default.

Command Modes

Router BGP neighbor configuration

Command History

Release

Modification

Release 4.2

This command was introduced.

Usage Guidelines

To configure BGP selective multipath feature, the multipath option must be enabled on the required BGP neighbor. The multipath configuration for a neighbor works when configured with the selective option of the maximum-paths command.

Task ID

Task ID

Operations

BGP

read, write

Examples

The following example shows how to enable multiple paths for a BGP neighbor.

RP/0/RSP0/CPU0:router# configure
RP/0/RSP0/CPU0:router(config)# router bgp 1
RP/0/RSP0/CPU0:router(config-bgp)# address-family ipv4 unicast
RP/0/RSP0/CPU0:router(config-bgp-af)# maximum-paths ibgp 4 selective
RP/0/RSP0/CPU0:router(config-bgp-af)# maximum-paths ebgp 5 selective
RP/0/RSP0/CPU0:router(config-bgp-af)# neighbor 1.1.1.1
RP/0/RSP0/CPU0:router(config-bgp-nbr)# address-family ipv4 unicast
RP/0/RSP0/CPU0:router(config-bgp-nbr-af)# multipath
RP/0/RSP0/CPU0:router(config-bgp-nbr-af)# commit

neighbor (BGP)

To enter neighbor configuration mode for configuring Border Gateway Protocol (BGP) routing sessions, use the neighbor command in an appropriate configuration mode. To delete all configuration for a neighbor and terminate peering sessions with the neighbor, use the no form of this command.

neighbor ip-address

no neighbor ip-address

Syntax Description

ip-address

IPv4 or IPv6 IP address of the BGP-speaking neighbor.

Command Default

Neighbor mode is not specified.

Command Modes

Router configuration

VRF configuration

Command History

Release

Modification

Release 3.7.2

This command was introduced.

Usage Guidelines

To use this command, you must be in a user group associated with a task group that includes appropriate task IDs. If the user group assignment is preventing you from using a command, contact your AAA administrator for assistance.

From router configuration mode, you can use this command to enter neighbor configuration mode.

From neighbor configuration mode, you can enter address family configuration for the neighbor by using the address-family command, which allows you to configure routing sessions for IP Version 4 and IP Version 6 address prefixes.

The neighbor command does not cause the neighbor to be configured and does not result in a peering to be established with the neighbor. To create the neighbor, you configure a remote autonomous system number by entering the remote-as command, or the neighbor can inherit a remote autonomous system from a neighbor group or session group if the use command is applied.


Note


A neighbor must have must a remote autonomous system number, and an IP address and address family must be enabled on the neighbor.


Unlike IPv4, IPv6 must be enabled before any IPv6 neighbors can be defined. Enable IPv6 in router configuration mode using the address-family command.


Note


Configuration for the neighbor cannot occur (peering is not established) until the neighbor is given a remote as-number and neighbor address.


The no form of this command causes the peering with the neighbor to be terminated and all configuration that relates to the neighbor to be removed.

Task ID

Task ID

Operations

bgp

read, write

Examples

The following example shows how to place the router in neighbor configuration mode for BGP routing process 1 and configure the neighbor IP address 172.168.40.24 as a BGP peer:

Router(config)# router bgp 1
Router(config-bgp)# neighbor 172.168.40.24
Router(config-bgp-nbr)# remote-as 65000

The following example shows how to enable IPv6 for BGP, then place the router in neighbor configuration mode for an IPv6 neighbor, 3000::1, and configure neighbor 3000::1 as a BGP peer:

Router(config)# router bgp 100 
Router(config-bgp)# address-family ipv6 unicast 
Router(config-bgp-af)# exit
Router(config-bgp)# neighbor 3000::1
Router(config-bgp-nbr)# remote-as 2002 
Router(config-bgp-nbr)# address-family ipv6 unicast 

neighbor-group

To create a neighbor group and enter neighbor group configuration mode, use the neighbor-group command in router configuration mode. To remove a neighbor group and delete all configuration associated with the group, use the no form of this command.

neighbor-group name

no neighbor-group name

Syntax Description

name

Neighbor group name.

Command Default

No neighbor group mode is specified.

Command Modes

Router configuration

Command History

Release

Modification

Release 3.7.2

This command was introduced.

Usage Guidelines

To use this command, you must be in a user group associated with a task group that includes appropriate task IDs. If the user group assignment is preventing you from using a command, contact your AAA administrator for assistance.

The neighbor-group command puts the router in neighbor group configuration mode and creates a neighbor group.

A neighbor group helps you apply the same configuration to one or more neighbors. After a neighbor group is configured, each neighbor can inherit the configuration through the use command. If a neighbor is configured to use a neighbor group, the neighbor, by default, inherits the entire configuration of the neighbor group, which includes the address family-independent and address family-specific configurations. The inherited configuration can be overridden if you directly configure commands for the neighbor or if you configure session groups or address family groups with the use command.

From neighbor group configuration mode, you can configure address family-independent parameters for the neighbor group. To enter address family-specific configuration for the neighbor group, use the address-family command when in the neighbor group configuration mode.


Note


If an address family is configured for a neighbor group, neighbors that use the neighbor group attempt to exchange routes in that address family.


The no form of this command ordinarily causes all configuration for the neighbor group to be removed. If using the no form would result in a neighbor losing its remote autonomous system number, the configuration is rejected. In this scenario, the neighbor configuration must be either removed or configured with a remote autonomous system number before the neighbor group configuration can be removed.


Note


Neighbor groups should not be configured with a mixture of IPv4 and IPv6 address families, because such a neighbor group is not usable by any neighbor. Note that within the Cisco IOS XR system configuration architecture, it is possible to create such a neighbor group; however, any attempt to use it is rejected.


Task ID

Task ID

Operations

bgp

read, write

Examples

The following example shows how to create a neighbor group called group1 that has IP Version 4 (IPv4) unicast and IPv4 multicast activated along with various configuration features. The neighbor group is used by neighbor 10.0.0.1 and neighbor 10.0.0.2, which allows them to inherit the entire group1 configuration.


Router(config)# router bgp 65530
Router(config-bgp)# neighbor-group group1
Router(config-bgp-nbrgrp)# remote-as 65535
Router(config-bgp-nbrgrp)# advertisement-interval 2
Router(config-bgp-nbrgrp)# address-family ipv4 unicast
Router(config-bgp-nbrgrp-af)# send-community-ebgp
Router(config-bgp-nbrgrp-af)# exit
Router(config-bgp-nbrgrp)# address-family ipv4 multicast
Router(config-bgp-nbrgrp-af)# next-hop-self
Router(config-bgp-nbrgrp-af)# exit
Router(config-bgp-nbrgrp)# exit
Router(config-bgp)#neighbor 10.0.0.1 
Router(config-bgp-nbr)# use neighbor-group group1
Router(config-bgp-nbr)# exit
Router(config-bgp)# neighbor 10.0.0.2
Router(config-bgp-nbr)# use neighbor-group group1
Router(config-bgp-nbr)# exit

neighbor internal-vpn-client

To preserve the iBGP-CE (customer edge) attributes inside the VPN attribute set (ATTR-SET) and send it across to the core, use the neighbor internal-vpn-client command in the VRF neighbor configuration mode. To disable the command, use the no form of this command.

neighbor ip-address internal-vpn-client

no neighbor ip-address internal-vpn-client

Syntax Description

neighbor ip-address

IP address of the neighboring device.

internal-vpn-client

Stacks the iBGP-CE neighbor path in the VPN attribute set.

Command Default

None

Command Modes

VRF neighbor configuration

Command History

Release

Modification

Release 5.3.1

This command was introduced.

Usage Guidelines

The neighbor ip-address internal-vpn-client command enables PE devices to make the entire VPN cloud act as an internal VPN client to the CE devices connected internally. This command is used so that existing internal BGP VRF lite scenarios are not affected. You need not configure autonomous system override for CE devices after enabling this command.

Task ID

Task ID

Operations

bgp

read, write

Examples

The following example shows how to configure L3VPN iBGP PE-CE:


RP/0/RSP0/CPU0:router# configure
RP/0/RSP0/CPU0:router(config)# router bgp 100
RP/0/RSP0/CPU0:router(config-bgp)# vrf blue neighbor 10.10.10.1
RP/0/RSP0/CPU0:router(config-bgp-vrf-nbr)# internal-vpn-client

network (BGP)

To specify that the Border Gateway Protocol (BGP) routing process should originate and advertise a locally known network to its neighbors, use the network command in an appropriate configuration mode. To disable originating or advertising the network to neighbors, use the no form of this command.

network {ip-address/prefix-length | | ip-address mask} [route-policy route-policy-name]

no network {ip-address/prefix-length | | ip-address mask} [route-policy route-policy-name]

Syntax Description

ip- address

Network that BGP advertises.

/ prefix-length

Length of the IP address prefix. A decimal value that indicates how many of the high-order contiguous bits of the address compose the prefix (the network portion of the address). A slash (/) must precede the decimal value.

ip-address mask

Network mask applied to the ip-address argument.

route-policy route-policy-name

(Optional) Specifies a route policy to use to modify the attributes of the network.

Command Default

No networks are specified.

Command Modes

IPv4 address family configuration

IPv6 address family configuration

VRF IPv4 address family configuration

VRF IPv6 address family configuration

Command History

Release

Modification

Release 3.7.2

This command was introduced.

Usage Guidelines

To use this command, you must be in a user group associated with a task group that includes appropriate task IDs. If the user group assignment is preventing you from using a command, contact your AAA administrator for assistance.

A network specified with this command is originated and advertised to neighbors only if there exists a route for the network in the routing table. That is, there must be a route learned using local or connected networks, static routing, or a dynamic IGP such as IS-IS or OSPF.

Other than the available system resources on the router, no limit exists on the number of network commands that can be configured.

Task ID

Task ID

Operations

bgp

read, write

Examples

The following example shows how to configure the local router to originate the IPv4 unicast network 172.20.0.0/16:

Router(config)#router bgp 120 
Router(config-bgp)# address-family ipv4 unicast
Router(config-bgp-af)# network 172.20.0.0/16

network backdoor

To set the administrative distance on an external Border Gateway Protocol (eBGP) route to that of a locally sourced BGP route, causing it to be less preferred than an Interior Gateway Protocol (IGP) route, use the network backdoor command in an appropriate configuration mode. To disable setting the administrative distance to the value for locally sourced BGP routes, use the no form of this command.

network {ip-address/prefix-length | | ip-address mask} backdoor

no network {ip-address/prefix-length | | ip-address mask} backdoor

Syntax Description

ip-address

Network that provides a backdoor route.

/ prefix-length

Length of the IP address prefix. A decimal value that indicates how many of the high-order contiguous bits of the address compose the prefix (the network portion of the address). A slash (/) must precede the decimal value.

mask

Network mask applied to the ip-address argument.

Command Default

No backdoor routes are installed.

Command Modes

IPv4 address family configuration

IPv6 address family configuration

VRF IPv4 address family configuration

VRF IPv6 address family configuration

Command History

Release

Modification

Release 3.7.2

This command was introduced.

Usage Guidelines

To use this command, you must be in a user group associated with a task group that includes appropriate task IDs. If the user group assignment is preventing you from using a command, contact your AAA administrator for assistance.

Configuring the network backdoor command does not cause BGP to originate a network, even if an IGP route for the network exists. Ordinarily, the backdoor network would be learned through both an eBGP and IGP. The BGP best-path selection algorithm does not change when a network is configured as a backdoor network.

Task ID

Task ID

Operations

bgp

read, write

Examples

The following example shows IP Version 4 (IPv4) unicast network 192.168.40.0/24 configured as a backdoor network:

Router(config)# router bgp 109
Router(config-bgp)# address-family ipv4 unicast
Router(config-bgp-af)# network 192.168.40.0/24 backdoor

next-hop-self

To disable next-hop calculation and insert your own address in the next-hop field of Border Gateway Protocol (BGP) updates, use the next-hop-self command in an appropriate configuration mode. To enable next-hop calculation, use the no form of this command.

next-hop-self [inheritance-disable]

no next-hop-self [inheritance-disable]

Syntax Description

inheritance-disable

(Optional) Allows a next-hop calculation override when this feature may be inherited from a neighbor group or address family group.

Command Default

When this command is not specified, the software calculates the next hop for BGP updates accepted by the router.

Command Modes

IPv4 address family group configuration

IPv6 address family group configuration

VPNv4 address family group configuration

IPv4 neighbor address family configuration

VPNv4 neighbor address family configuration

IPv4 neighbor group address family configuration

IPv6 neighbor group address family configuration

VPNv4 neighbor group address family configuration

Command History

Release

Modification

Release 3.7.2

This command was introduced.

Release 3.9.0

The disable keyword was replaced with the inheritance-disable keyword.

Usage Guidelines

To use this command, you must be in a user group associated with a task group that includes appropriate task IDs. If the user group assignment is preventing you from using a command, contact your AAA administrator for assistance.

Use the next-hop-self command to set the BGP next-hop attribute of routes being advertised over a peering session to the local source address of the session.

This command is useful in nonmeshed networks in which BGP neighbors may not have direct access to all other neighbors on the same IP subnet.

If this command is configured for a neighbor group or address family group, a neighbor using the group inherits the configuration. Configuring the command specifically for a neighbor overrides any inherited value.

Configuring the next-hop-self command under IPv4 labeled-unicast, IPv6 labeled-unicast, or VRF labeled-unicast address family configuration mode enables next-hop-self for labeled prefixes advertised to an iBGP peer.

Task ID

Task ID

Operations

bgp

read, write

Examples

The following example shows how to set the next hop of the update field for all IP Version 4 (IPv4) unicast routes advertised to neighbor 172.20.1.1 to an address of the local router:

RP/0/RSP0/CPU0:router(config)# router bgp 140
RP/0/RSP0/CPU0:router(config-bgp)# neighbor 172.20.1.1
RP/0/RSP0/CPU0:router(config-bgp-nbr)# remote-as 1
RP/0/RSP0/CPU0:router(config-bgp-nbr)# address-family ipv4 unicast
RP/0/RSP0/CPU0:router(config-bgp-nbr-af)# next-hop-self

The following example shows how to disable the next-hop-self command for neighbor 172.20.1.1. If not overridden, the next hop would be inherited from address family group group1:

RP/0/RSP0/CPU0:router(config)# router bgp 140
RP/0/RSP0/CPU0:router(config-bgp)# af-group group1 address-family ipv4 unicast
RP/0/RSP0/CPU0:router(config-bgp-afgrp)# next-hop-self
RP/0/RSP0/CPU0:router(config-bgp-afgrp)# exit
RP/0/RSP0/CPU0:router(config-bgp)# neighbor 172.20.1.1
RP/0/RSP0/CPU0:router(config-bgp-nbr)# remote-as 1
RP/0/RSP0/CPU0:router(config-bgp-nbr)# address-family ipv4 unicast
RP/0/RSP0/CPU0:router(config-bgp-nbr-af)# use af-group group1
RP/0/RSP0/CPU0:router(config-bgp-nbr-af)# next-hop-self inheritance-disable

next-hop-unchanged

To disable overwriting of the next hop before advertising to external Border Gateway Protocol (eBGP) peers, use the next-hop-unchanged command in an appropriate configuration mode. To enable overwriting of the next hop, use the no form of this command.

next-hop-unchanged [inheritance-disable]

no next-hop-unchanged [inheritance-disable]

Syntax Description

inheritance-disable

(Optional) Allows overwriting of the next hop before advertising to eBGP peers when this feature may be inherited from a neighbor group or address family group.

Command Default

Overwriting of the next hop is allowed.

Command Modes

VPNv4 address family group configuration

VPNv4 neighbor address family configuration

VPNv4 neighbor group address family configuration

IPv4 labeled-unicast address family configuration

IPv6 labeled-unicast address family configuration

IPv4 address family configuration

IPv6 address family configuration

Command History

Release

Modification

Release 3.7.2

This command was introduced.

Release 3.9.0

The disable keyword was replaced with the inheritance-disable keyword.

Release 4.0.0

This command was supported in the following address family configuration modes:

  • IPv4 labeled-unicast address family configuration

  • IPv6 labeled-unicast address family configuration

  • IPv4 unicast address family configuration

  • IPv6 unicast address family configuration

Usage Guidelines

Use the next-hop-unchanged command to propagate the next hop unchanged for multihop eBGP peering sessions. This command should not be configured on a route reflector, and the next-hop-self command should not be used to modify the next-hop attribute for a route reflector when this feature is enabled for a route reflector client.


Note


Incorrectly setting BGP attributes for a route reflector can cause inconsistent routing, routing loops, or a loss of connectivity. Setting BGP attributes for a route reflector should be attempted only by an experienced network operator.

Task ID

Task ID

Operations

bgp

read, write

Examples

The following example shows how to disable the overwriting of next hops before advertising to eBGP peers:

RP/0/RSP0/CPU0:router(config)# router bgp 140
RP/0/RSP0/CPU0:router(config-bgp)# af-group group1 address-family ipv4 unicast
RP/0/RSP0/CPU0:router(config-bgp-afgrp)# next-hop-unchanged disable
RP/0/RSP0/CPU0:router(config-bgp-afgrp)# exit

nexthop resolution prefix-length minimum

To set minimum prefix-length for nexthop resolution, use the nexthop resolution prefix-length minimum command in an appropriate configuration mode. To disable the minimum prefix-length for nexthop resolution, use the no form of this command.

nexthop resolution prefix-length minimum prefix-length-value

no nexthop resolution prefix-length minimum prefix-length-value

Syntax Description

prefix-length-value

Sets the minimum prefix-length. Range is 0 to 32 in VPNv4 and VRF IPv4 Unicast. Range is 0 to 128 in IPv6 Unicast.

Command Default

Nexthop resolution for minimum prefix-length is disabled.

Command Modes

VPNv4 Unicast address family

VPNv6 Unicast address family

IPv4 Unicast address family

IPv6 Unicast address family

Link-state address family

L2VPN VPLS-VPWS address family

L2VPN EVPN address family

L2VPN MSPW address family

Command History

Release Modification

Release 7.0.2, 7.2.1

This command was supported under IPv6 Unicast address family.

Release 4.2.1

This command was introduced.

Usage Guidelines

To use this command, you must be in a user group associated with a task group that includes appropriate task IDs. If the user group assignment is preventing you from using a command, contact your AAA administrator for assistance.

Task ID

Task ID Operation

bgp

read, write

Examples

This example shows how to set the minimum prefix-length for nexthop resolution as 32 :


Router#configure
Router(config)#router bgp 100
Router(config-bgp)#address-family vpnv4 unicast
Router(config-bgp-af)#nexthop resolution prefix-length minimum 32

nexthop route-policy

To specify that BGP routes are resolved using only next hops whose routes match specific characteristics, use the nexthop route-policy command in the appropriate configuration mode. To remove the nexthop route-policy command from the configuration file and restore the system to its default behavior, use the no form of this command.

nexthop route-policy route-policy-name

no nexthop route-policy route-policy-name

Syntax Description

route-policy-name

Route policy to use for filtering based on next hops.

Command Default

No default behavior or values

Command Modes

IPv4 address family configuration

IPv6 address family configuration

VPNv4 address family configuration

VRF address family configuration

Command History

Release

Modification

Release 3.7.2

This command was introduced.

Release 7.11.1

This command was modified. This command is extended to VRF address-family configuration mode.

Usage Guidelines

To use this command, you must be in a user group associated with a task group that includes appropriate task IDs. If the user group assignment is preventing you from using a command, contact your AAA administrator for assistance.

Use the nexthop route-policy command to configure route policy filtering using next hops.

The BGP next-hop tracking feature allows you to specify that BGP routes are resolved using only next hops whose routes have the following characteristics:

  • To avoid the aggregate routes, the prefix length must be greater than a specified value.
  • The source protocol must be from a selected list, ensuring that BGP routes are not used to resolve next hops that could lead to oscillation.

This route policy filtering is possible because RIB identifies the source protocol of a route that resolves a next hop as well as the mask length associated with the route.

The next-hop attach point supports matching using the protocol name and mask length. BGP marks all next hops that are rejected by the route policy as invalid, and no best path is calculated for the routes that use the invalid next hop. The invalid next hops continue to stay in the active cache and can be displayed as part of the show bgp nexthop command with an invalid status.

Task ID

Task ID

Operations

bgp

read, write

Examples

The following example shows how to specify the route policy nexthop_A as the policy to use for filtering next hops:


Router(config)# router bgp 109
Router(config-bgp)# address-family ipv4 unicast 
Router(config-bgp-af)# nexthop route-policy nexthop_A

Examples

The following example shows how to specify the route policy nexthop_A as the policy to use for filtering next hops in the VRF address family configuration mode:

Router(config)# router bgp 500
Router(config-bgp)# vrf vrf10 
Router(config-bgp-vrf)# address-family ipv4 unicast
Router(config-bgp-vrf-af)# nexthop route-policy nh-route-policy

nexthop trigger-delay

To specify the delay for triggering next-hop calculations, use the nexthop trigger-delay command in the appropriate configuration mode. To set the trigger delay to the default value, use the no form of this command.

nexthop trigger-delay {critical delay | non-critical delay}

no nexthop trigger-delay {critical delay | non-critical delay}

Syntax Description

critical

Specifies critical next-hop events. For example, when the next hop is unreachable.

delay

Trigger delay, in milliseconds. Range is 0 to 4294967295.

non-critical

Specifies noncritical next-hop events. For example, Interior Gateway Protocol (IGP) metric changes.

Command Default

critical : 3000 msec for all address families, except VPNv4 and VPNv6 address families

critical : 50 msec for VPNv4 and VPNv6 address families

non-critical : 10000 msec for all address families

Command Modes

All address families

Command History

Release

Modification

Release 3.7.2

This command was introduced.

Release 7.10.1

The default value of critical delay for VPN address family was changed from 0 msec to 50 msec.

Usage Guidelines

To use this command, you must be in a user group associated with a task group that includes appropriate task IDs. If the user group assignment is preventing you from using a command, contact your AAA administrator for assistance.

Use the nexthop trigger-delay command to allow for a dynamic way for Interior Gateway Protocol (IGP) to converge. This convergence allows BGP to accumulate all notifications and trigger fewer walks, resulting in fewer interprocess communications (IPCs) to the Routing Information Base (RIB) for route addition, deletion, and modification and fewer updates to peers.


Note


A high delay value can be configured to effectively turn off next-hop tracking.


The non-critical delay value must always be set to at least equal or greater than the critical delay value .

The delay should be slightly higher than the time it takes for the IGP to settle into a steady state after some event (IGP convergence time).

Avoid configuring the nexthop trigger-delay critical 0 as it is not suitable on:

  • Scaled deployments where a long BGP next-hop walk time duration is inevitable.

  • Deployments where BGP next-hop changes are frequent.

Disadvantages of nexthop trigger-delay critical 0 configuration

  • High CPU utilization as each change notification triggers a BGP next-hop walk for address families configured with nexthop trigger-delay critical 0.

  • BGP next-hop change notifications are not batched. This disallows interleaving of next-hop walks in address families with the non-zero delay configuration as these address families wait until the address families with the zero critical delay value complete their next-hop walks.

  • Extended wait time before the BGP next-hop walk starts on address families with the non-zero critical delay configuration, leading to potential traffic blackholing.

Starting with Cisco IOS XR Release 7.10.1, the default critical delay configuration in VPNv4 address family was changed from 0 msec to 50 msec. With this change, all address families have a default non-zero critical delay value. To see the critical delay value of each address family, run the show bgp all all nexthops command.

After you have upgraded to Cisco IOS XR Release 7.10.1 or later, if you configure the default critical delay value in the IPv4 address family to 0 msec, you will observe a considerable delay in VPNv4 convergence for the following reasons:

  • The IPv4 address families are walked as many times as the number of next-hop critical alerts raised to BGP.

  • The BGP next-hop updates for the IPv4 address family prefixes take precedence over VPNv4 address family prefixes.

Advantages of configuring nexthop trigger-delay critical with a non-zero default value

  • Provides next-hop change notification batching which reduces the number of BGP next-hop walks.

  • Allows interleaving different active BGP next-hop walks for the respective address families while prioritizing some address families over the others.

Therefore, we strongly recommend you to configure nexthop trigger-delay critical with a non-zero value.

Task ID

Task ID

Operations

bgp

read, write

Examples

This example shows how to set the critical next-hop trigger delay to 3500 milliseconds:

RP/0/RSP0/CPU0:router(config)# router bgp 109
RP/0/RSP0/CPU0:router(config-bgp)# address-family ipv4 unicast 
RP/0/RSP0/CPU0:router(config-bgp-af)# nexthop trigger-delay critical 3500 

nsr (BGP)

To activate Border Gateway Protocol (BGP) nonstop routing (NSR), use the nsr command in BGP global configuration mode. To deactivate BGP NSR, use the no form of this command.

nsr

no nsr

Syntax Description

This command has no arguments or keywords.

Command Default

BGP NSR is not activated.

Command Modes

Global configuration

Command History

Release

Modification

Release 3.9.0

This command was introduced.

Usage Guidelines

To use this command, you must be in a user group associated with a task group that includes appropriate task IDs. If the user group assignment is preventing you from using a command, contact your AAA administrator for assistance.

Use the nsr command to enable the Border Gateway Protocol (BGP) Nonstop Routing (NSR) with Stateful Switchover (SSO). This enables all bgp peerings to maintain the BGP state to ensure continuous packet forwarding during events that could interrupt service.

Note


From release 5.2.3, NSR is enabled by default.


Task ID

Task ID

Operations

bgp

read, write

Examples

The following example shows how to enable BGP NSR:


  RP/0/RSP0/CPU0:router# configure 
  RP/0/RSP0/CPU0:router(config)# router bgp 120 
  RP/0/RSP0/CPU0:router(config-bgp)# nsr 
  

The following example shows how to disable BGP NSR:


  RP/0/RSP0/CPU0:router# configure
  RP/0/RSP0/CPU0:router(config)# router bgp 120
  RP/0/RSP0/CPU0:router(config-bgp)# no nsr          
  

nsr disable (BGP)

To disable Border Gateway Protocol (BGP) nonstop routing (NSR), use the nsr disable command in BGP global configuration mode. To re-enable BGP NSR, use the no form of this command.

nsr disable

no nsr disable

Syntax Description

This command has no arguments or keywords.

Command Default

BGP NSR is activated by default.

Command Modes

Global configuration

Command History

Release

Modification

Release 5.3.0

This command was introduced.

Usage Guidelines

Use the nsr disable command to disable Border Gateway Protocol (BGP) Nonstop Routing (NSR) with Stateful Switchover (SSO). Using the no form of this command enables all BGP peerings to maintain the BGP state to ensure continuous packet forwarding during events that could interrupt service.

Note


In releases prior to R 5.2.3, NSR is disabled by default, and must be configured manually.


Task ID

Task ID

Operations

bgp

read, write

Examples

The following example shows how to disable BGP NSR:


  RP/0/RSP0/CPU0:router# configure 
  RP/0/RSP0/CPU0:router(config)# router bgp 120 
  RP/0/RSP0/CPU0:router(config-bgp)# nsr disable
  

The following example shows how to re-enable BGP NSR:


  RP/0/RSP0/CPU0:router# configure
  RP/0/RSP0/CPU0:router(config)# router bgp 120
  RP/0/RSP0/CPU0:router(config-bgp)# no nsr disable          
  

orf

To specify Outbound Route Filter (ORF) and inbound filtering criteria, use the orf route-policy command in an appropriate configuration mode. To restore the system to its default condition, use the no form of this command.

orf route-policy route-policy-name

no orf route-policy route-policy-name

Syntax Description

route-policy-name

Name of the route policy.

Command Default

No ORF route policy is defined.

Command Modes

IPv4 address family group configuration

IPv6 address family group configuration

IPv4 neighbor address family configuration

IPv4 neighbor group address family configuration

IPv6 neighbor group address family configuration

VRF IPv4 neighbor address family configuration

VRF IPv6 neighbor address family configuration

Command History

Release

Modification

Release 3.7.2

This command was introduced.

Usage Guidelines

To use this command, you must be in a user group associated with a task group that includes appropriate task IDs. If the user group assignment is preventing you from using a command, contact your AAA administrator for assistance.

Task ID

Task ID

Operations

bgp

read, write

Examples

The following example shows how to configure outbound and inbound filtering criteria:


RP/0/RSP0/CPU0:router(config)#router bgp 6 
RP/0/RSP0/CPU0:router(config-bgp)# neighbor 172.20.1.1
RP/0/RSP0/CPU0:router(config-bgp-nbr)# remote-as 1
RP/0/RSP0/CPU0:router(config-bgp-nbr)# address-family ipv4 unicast
RP/0/RSP0/CPU0:router(config-bgp-nbr-af)#orf route-policy policy_A  

password (BGP)

To enable Message Digest 5 (MD5) authentication on a TCP connection between two Border Gateway Protocol (BGP) neighbors, use the password command in an appropriate configuration mode. To disable MD5 authentication, use the no form of this command.

password {clear | encrypted} password

no password [clear password | encrypted password]

Syntax Description

clear

Specifies that an unencrypted password follows. The password must be a case-sensitive, clear-text unencrypted password.

encrypted

Specifies that an encrypted password follows. The password must be a case-sensitive, encrypted password.

password

Password of up to 80 characters. The password can contain any alphanumeric characters. However, if the first character is a number or the password contains a space, the password must be enclosed in double quotation marks; for example, “2 password.”

Command Default

When this command is not specified in the appropriate configuration mode, MD5 authentication is not enabled on a TCP connection between two BGP neighbors.

Command Modes

Neighbor configuration

VRF neighbor configuration

Neighbor group configuration

Session group configuration

Command History

Release

Modification

Release 3.7.2

This command was introduced.

Usage Guidelines

To use this command, you must be in a user group associated with a task group that includes appropriate task IDs. If the user group assignment is preventing you from using a command, contact your AAA administrator for assistance.

Configure a password to enable authentication between two BGP peers. Use the password command to verify each segment sent on the TCP connection between the peers. The same password must be configured on both networking devices, otherwise a connection cannot be made. The authentication feature uses the MD5 algorithm. Specifying this command causes the software to generate and check the MD5 digest on every segment sent on the TCP connection.

Configuring a neighbor password does not cause the existing session for a neighbor to end. However, until the new password is configured on the remote router, the local BGP process does not receive keepalive messages from the remote device. If the password is not updated on the remote device by the end of the hold time, the session ends. The hold time can be changed using the timers command or the timers bgp command.

If this command is configured for a neighbor group or neighbor address family group, a neighbor using the group inherits the configuration. Values of commands configured specifically for a neighbor overrides inherited values.

Task ID

Task ID

Operations

bgp

read, write

Examples

The following example shows how to configure neighbor 172.20.1.1 to use MD5 authentication with the password password1:


RP/0/RSP0/CPU0:router(config)# router bgp 140
RP/0/RSP0/CPU0:router(config-bgp)#neighbor 172.20.1.1 
RP/0/RSP0/CPU0:router(config-bgp-nbr)#remote-as 1 
RP/0/RSP0/CPU0:router(config-bgp-nbr)#password clear password1 

password (rpki-server)

To specify a SSH password for the RPKI cache-server, use the password command in rpki-server configuration mode. To remove the SSH passwords, use the no form of this command.

password password

no password password

Syntax Description

password

Enters a password to be used for the SSH transport mechanism.

Command Default

Password is not configured.

Command Modes

RPKI server configuration

Command History

Release Modification

Release 4.2.1

This command was introduced.

Usage Guidelines

To use this command, you must be in a user group associated with a task group that includes appropriate task IDs. If the user group assignment is preventing you from using a command, contact your AAA administrator for assistance.

SSH expects to use an authentication method to connect to a remote server. The SSH authentication method to connect to RPKI server is password-based. So, the RPKI cache-server must be configured with username and password. A username and password must be configure for each server configured under BGP that uses the SSH transport

Task ID

Task ID Operation

bgp

read, write

Examples

This example shows how to configure a username (rpki-user ) and password (rpki-ssh-pass ) for the RPKI cache-server SSH transport mechanism:


Router#configure
Router(config)#router bgp 100
Router(config-bgp)#rpki server 172.168.35.40 
Router(config-bgp-rpki-server)# transport ssh port 22
Router(config-bgp-rpki-server)#username rpki-user
Router(config-bgp-rpki-server)#password rpki-ssh-pass

password-disable

To override any inherited password configuration from a neighbor group or session group for Border Gateway Protocol (BGP) neighbors, use the password-disable command in an appropriate configuration mode. To disable overriding any inherited password command, use the no form of this command.

password-disable

no password-disable

Syntax Description

This command has no arguments or keywords.

Command Default

Configured passwords for neighbor and session groups are inherited.

Command Modes

Neighbor configuration

VRF neighbor configuration

Neighbor group configuration

Session group configuration

Command History

Release

Modification

Release 3.7.2

This command was introduced.

Usage Guidelines

To use this command, you must be in a user group associated with a task group that includes appropriate task IDs. If the user group assignment is preventing you from using a command, contact your AAA administrator for assistance.

If you specify a password on a neighbor group or session group, all users of the group inherit the password. Specifying a different password command specifically on a neighbor that uses the group overrides the inherited value. Specifying password-disable on a neighbor that uses the group disables password authentication for the neighbor.

Task ID

Task ID

Operations

bgp

read, write

Examples

The following example shows how to disable MD5 authentication for neighbor 172.20.1.1, preventing it from inheriting the password password1 from session group group1:


RP/0/RSP0/CPU0:router(config)# router bgp 140
RP/0/RSP0/CPU0:router(config-bgp)# session-group group1
RP/0/RSP0/CPU0:router(config-bgp-sngrp)# password clear password1
RP/0/RSP0/CPU0:router(config-bgp-sngrp)# exit
RP/0/RSP0/CPU0:router(config-bgp)# neighbor 172.20.1.1
RP/0/RSP0/CPU0:router(config-bgp-nbr)# remote-as 2
RP/0/RSP0/CPU0:router(config-bgp-nbr)# use session-group group1
RP/0/RSP0/CPU0:router(config-bgp-nbr)# password-disable

permanent-network

To define a prefix set as permanent, use the permanent-network command in the global address family configuration mode. To remove a prefix set as permanent, use the no form of this command. The permanent-network command uses a route-policy to identify the set of prefixes (networks) for which permanent paths needs to be created.

The permanent network feature supports only prefixes in IPv4 unicast and IPv6 unicast address-families under the default Virtual Routing and Forwarding (VRF).

permanent-network route-policy route-policy-name

no permanent-network

Syntax Description

route-policy route-policy-name

Specifies a configured routing policy.

Command Default

None

Command Modes

Address-family configuration.

Command History

Release

Modification

Release 5.1.1

This command was introduced.

Usage Guidelines

To use this command, you must be in a user group associated with a task group that includes appropriate task IDs. If the user group assignment is preventing you from using a command, contact your AAA administrator for assistance.

Task ID

Task ID

Operations

bgp

read, write

Examples

This example shows how to define permanent path for a route policy named POLICY-PERMANENT-NETWORK-IPv4:


RP/0/RSP0/CPU0:router# configure
RP/0/RSP0/CPU0:router(config)# router bgp 100
RP/0/RSP0/CPU0:router(config-bgp)# address-family ipv4 unicast
RP/0/RSP0/CPU0:router(config-af)# permanent-network route-policy POLICY-PERMANENT-NETWORK-IPv4

precedence

To set the precedence level, use the precedence command in the appropriate configuration mode. To remove the precedence command from the configuration file and restore the system to its default interval values, use the no form of this command.

precedence value

no precedence [value]

Syntax Description

value

Value of the precedence. The precedence value can be a number from 0 to 7, or it can be one of the following keywords:

critical —Set packets with critical precedence (5)

flash — Set packets with flash precedence (3)

flash-override —Set packets with flash override precedence (4)

immediate —Set packets with immediate precedence (2)

internet —Set packets with internetwork control precedence (6)

network —Set packets with network control precedence (7)

priority —Set packets with priority precedence (1)

routine —Set packets with routine precedence (0)

Command Default

No default behavior or values

Command Modes

Neighbor configuration

Neighbor session group configuration

Neighbor group configuration

Command History

Release

Modification

Release 3.7.2

This command was introduced.

Usage Guidelines

To use this command, you must be in a user group associated with a task group that includes appropriate task IDs. If the user group assignment is preventing you from using a command, contact your AAA administrator for assistance.

Use the precedence command to set the precedence value.

Task ID

Task ID

Operations

bgp

read, write

Examples

The following example shows how to set the precedence to 2:

RP/0/RSP0/CPU0:router(config)# router bgp 5
RP/0/RSP0/CPU0:router(config-bgp)# neighbor 10.1.1.1 
RP/0/RSP0/CPU0:router(config-bgp-nbr)# remote-as 100
RP/0/RSP0/CPU0:router(config-bgp-nbr)# precedence 2 

preference (rpki-server)

To specify a preference value for the RPKI cache-server, use the preference command rpki-server configuration mode. To remove the preference value, use the no form of this command.

preference preference-value

no preference preference-value

Syntax Description

preference-value

Specifies a RPKI cache preference value. Range is 1 to 10.

Note

 
A lower value is recommended

Command Default

Preference value is not set.

Command Modes

RPKI server configuration

Command History

Release Modification

Release 4.2.1

This command was introduced.

Usage Guidelines

To use this command, you must be in a user group associated with a task group that includes appropriate task IDs. If the user group assignment is preventing you from using a command, contact your AAA administrator for assistance.

Task ID

Task ID Operation

bgp

read, write

Examples

This example shows how to set preference value for RPKI configuration as 1:


Router#configure
Router(config)#router bgp 100
Router(config-bgp)#rpki server 172.168.35.40 
Router(config-bgp-rpki-cache)# transport ssh port 22
Router(config-bgp-rpki-cache)#username rpki-user
Router(config-bgp-rpki-cache)#password rpki-ssh-pass
Router(config-bgp-rpki-cache)#preference 1

purge-time (rpki-server)

To configure the time BGP waits to keep routes from RPKI cache-server after the cache session drops, use the purge-time command in rpki-server configuration mode. To remove the purge-time configuration, use the no form of this command.

purge-time time-in-seconds

no purge-time time-in-seconds

Syntax Description

time-in-seconds

Sets the purge time in seconds. Range is 30 to 360 seconds.

Command Default

Sets a default value of 60 seconds.

Command Modes

RPKI server configuration

Command History

Release Modification

Release 4.2.1

This command was introduced.

Usage Guidelines

To use this command, you must be in a user group associated with a task group that includes appropriate task IDs. If the user group assignment is preventing you from using a command, contact your AAA administrator for assistance.

Task ID

Task ID Operation

bgp

read, write

Examples

This example shows how to set the purge-time for RPKI cache as 30 seconds:


Router#configure
Router(config)#router bgp 100
Router(config-bgp)#rpki server 172.168.35.40 
Router(config-bgp-rpki-server)# transport ssh port 22
Router(config-bgp-rpki-server)#username rpki-user
Router(config-bgp-rpki-server)#password rpki-ssh-pass
Router(config-bgp-rpki-server)#preference 1
Router(config-bgp-rpki-server)#purge-time 30

rd

To configure a route distinguisher, use the rd command in VRF configuration mode. To disable the route distinguisher, use the no form of this command.

rd {as-number : nn | ip-address : nn | auto}

no rd {as-number : nn | ip-address : nn | auto}

Syntax Description

as-number:nn

  • as-number —16-bit Autonomous system (AS) number of the route distinguisher
    • Range for 2-byte Autonomous system numbers (ASNs) is 1 to 65535.

    • Range for 4-byte Autonomous system numbers (ASNs) in asplain format is 1 to 4294967295.

    • Range for 4-byte Autonomous system numbers (ASNs) is asdot format is 1.0 to 65535.65535.

  • nn —32-bit number

ip-address:nn

IP address of the route distinguisher.

  • ip-address —32-bit IP address
  • nn —16-bit number

auto

Automatically assigns a unique route distinguisher.

Command Default

No default behavior or values

Command Modes

VRF configuration

Command History

Release

Modification

Release 3.7.2

This command was introduced.

Release 3.9.0

Asplain format for 4-byte Autonomous system numbers notation was supported.

Usage Guidelines

To use this command, you must be in a user group associated with a task group that includes appropriate task IDs. If the user group assignment is preventing you from using a command, contact your AAA administrator for assistance.

Use the rd command to make the prefix unique across multiple VRFs.

Auto assignment of route distinguishers can be done only if a router ID is assigned using the bgp router-id command in BGP router configuration mode. The unique router ID is used for automatic route distinguisher generation.

The following are restrictions when configuring route distinguishers:

  • BGP router-id must be configured before rd auto can be configured
  • Route distinguisher cannot be changed or removed when an IPv4 unicast address family is configured under VRF.
  • BGP router-id cannot be changed or removed when rd auto is configured under a VRF.
  • When rd auto is configured under a VRF, the IP address for the router distinguisher configured under another VRF must be different from that of the BGP router-id
  • If a route distinguisher with same IP address as BGP router-id exists, the rd auto is not permitted.

Task ID

Task ID

Operations

bgp

read, write

Examples

The following example shows how to automatically assign a unique route distinguisher to VRF instance vrf-1:


RP/0/RSP0/CPU0:router(config)# router bgp 1
RP/0/RSP0/CPU0:router(config-bgp)# vrf vrf-1
RP/0/RSP0/CPU0:router(config-bgp-vrf)# rd auto

receive-buffer-size

To set the size of the receive buffers for a Border Gateway Protocol (BGP) neighbor, use the receive-buffer-size command in an appropriate configuration mode. To remove the receive-buffer-size command from the configuration file and restore the system to its default condition in which the software uses the default size, use the no form of this command.

receive-buffer-size socket-size [bgp-size]

no receive-buffer-size [socket-size] [bgp-size]

Syntax Description

socket-size

Size, in bytes, of the receive-side socket buffer. Range is 512 to 131072.

bgp-size

(Optional) Size, in bytes, of the receive buffer in BGP. Range is 512 to 131072.

Command Default

socket-size : 32,768 bytes

bgp-size : 4,032 bytes

Command Modes

Neighbor configuration

VRF neighbor configuration

Neighbor group configuration

Session group configuration

Command History

Release

Modification

Release 3.7.2

This command was introduced.

Usage Guidelines

To use this command, you must be in a user group associated with a task group that includes appropriate task IDs. If the user group assignment is preventing you from using a command, contact your AAA administrator for assistance.

Use the receive-buffer-size command to increase the buffer size when receiving updates from a neighbor. Using larger buffers can improve convergence time because it allows the software to process a larger number of packets simultaneously. However, allocating larger buffers consumes more memory on the router.


Note


Increasing the socket buffer size uses more memory only when more messages are waiting to be processed by the software. In contrast, increasing the BGP buffer size uses extra memory indefinitely.


If this command is configured for a neighbor group or session group, all neighbors using the group inherit the configuration. Values of commands configured specifically for a neighbor override inherited values.

Task ID

Task ID

Operations

bgp

read, write

Examples

The following example shows how to set the receive buffer sizes for neighbor 172.20.1.1 to be 65,536 bytes for the socket buffer and 8192 bytes for the BGP buffer:

RP/0/RSP0/CPU0:router(config)# router bgp 1
RP/0/RSP0/CPU0:router(config-bgp)# neighbor 172.20.1.1
RP/0/RSP0/CPU0:router(config-bgp-nbr)# remote-as 1
RP/0/RSP0/CPU0:router(config-bgp-nbr)# receive-buffer-size 65536 8192

redistribute (BGP)

To redistribute routes from one routing domain into Border Gateway Protocol (BGP), use the redistribute command in an appropriate configuration mode. To disable route redistribution, use the no form of this command.

Connected

redistribute connected [metric metric-value] [route-policy route-policy-name]

no redistribute connected [metric metric-value] [route-policy route-policy-name]

Enhanced Interior Gateway Routing Protocol (EIGRP)

redistribute eigrp process-id [match {external | internal}] [metric metric-value] [route-policy route-policy-name]

no redistribute eigrp process-id [match {external | internal}] [metric metric-value] [route-policy route-policy-name]

Intermediate System-to-Intermediate System (IS-IS)

redistribute isis process-id [level | | {1 | | 1-inter-area | | 2}] [metric metric-value] [route-policy route-policy-name]

no redistribute isis process-id [level | | {1 | | 1-inter-area | | 2}] [metric metric-value] [route-policy route-policy-name]

Open Shortest Path First (OSPF)

redistribute ospf process-id

no redistribute ospf process-id

Routing Information Protocol

redistribute rip [metric metric-value] [route-policy route-policy-name]

no redistribute rip [metric metric-value] [route-policy route-policy-name]

Static

redistribute static [metric metric-value] [route-policy route-policy-name]

no redistribute static [metric metric-value] [route-policy route-policy-name]

Syntax Description

connected

Redistributes connected routes. Connected routes are established automatically when IP is enabled on an interface.

metric metric-value

(Optional) Specifies the Multi Exit Discriminator (MED) attribute used for the redistributed route. Range is 0 to 4294967295. Use a value consistent with the destination protocol.

By default, the Interior Gateway Protocol (IGP) metric is assigned to the route. For connected and static routes the default metric is 0.

route-policy route-policy-name

(Optional) Specifies a configured routing policy to filter redistributed routes. A route policy is used to filter the importation of routes from this source routing protocol to BGP.

eigrp

Specifies that routes are distributed from EIGRP. You must be in IPv4 unicast or multicast address family configuration mode or in VRF IPv4 address family configuration mode.

process-id

For the eigrp keyword, an EIGRP instance name from which routes are to be redistributed.

For the isis keyword, an IS-IS instance name from which routes are to be redistributed.

For the ospf keyword, an OSPF instance name from which routes are to be redistributed.

The process-id value takes the form of a string. A decimal number can be entered, but it is stored internally as a string.

match { internal | external [ 1 | 2 ] | nssa-external [ 1 | 2 ]}

(Optional) Specifies the criteria by which OSPF routes are redistributed into other routing domains. It can be one or more of the following:

  • internal —Routes that are internal to a specific autonomous system (intra- and inter-area OSPF routes).
  • external [ 1 | 2 ]—Routes that are external to the autonomous system, but are imported into OSPF as Type 1 or Type 2 external routes.
  • nssa-external [ 1 | 2 ]—Routes that are external to the autonomous system, but are imported into OSPF as Type 1 or Type 2 not-so-stubby area (NSSA) external routes.

For the external and nssa-external options, if a type is not specified, then both Type 1 and Type 2 are assumed.

isis

Specifies that routes are distributed from the IS-IS protocol.

Redistribution from IS-IS is allowed under IPv4 unicast, IPv4 multicast, IPv6 unicast, and address-families. Redistribution is not allowed under VPNv4 address-families.

level { 1 | 1-inter-area | 2 }

(Optional) Specifies the IS-IS level from which routes are redistributed. It can be one of the following:

  • 1 —Routes are redistributed from Level 1 routes.
  • 1-inter-area —Routes are redistributed from Level 1 interarea routes.
  • 2 —Routes are redistributed from Level 2 routes.

ospf

Specifies that routes are distributed from the OSPF protocol. You must be in IPv4 unicast or multicast address family configuration mode or in VRF IPv4 address family configuration mode.

rip

Specifies that routes are distributed from RIP. You must be in IPv4 unicast or multicast address family configuration mode.

static

Redistributes IP static routes.

Command Default

Route redistribution is disabled.

For IS-IS, the default is to redistribute Level 1 and Level 2 routes.

For OSPF, the default is to redistribute internal, external, and NSSA external routes of Type 1 and Type 2.

By default, the Interior Gateway Protocol (IGP) metric is assigned to the route. For connected and static routes the default metric is 0.

metric metric-value: 0

match { internal | external [ 1 | 2 ] | nssa-external [ 1 | 2 ]}: If no match is specified, the default is to match all routes.

Command Modes

IPv4 address family configuration, both unicast and multicast (connected, eigrp, isis, ospf, rip, and static are supported)


            IPv6 address family configuration, both unicast and multicast (connected, eigrp,
               isis,
            

            ospfv3,
            and static are supported)
        

VRF IPv4 address family configuration ( connected , eigrp , ospf , rip , and static are supported)

VRF IPv6 address family configuration ( connected , eigrp , and static are supported)

Command History

Release

Modification

Release 3.7.2

This command was introduced.

Usage Guidelines

To use this command, you must be in a user group associated with a task group that includes appropriate task IDs. If the user group assignment is preventing you from using a command, contact your AAA administrator for assistance.


Note


When redistributing routes (into BGP) using both command keywords for setting or matching of attributes and a route policy, the routes are run through the route policy first, followed by the keyword matching and setting.


Each instance of a protocol may be redistributed independently of the others. Changing or removing redistribution for a particular instance does not affect the redistribution capability of other protocols or other instances of the same protocol.

Networks specified using the network command are not affected by the redistribute command; that is, the routing policy specified in the network command takes precedence over the policy specified through the redistribute command.

Task ID

Task ID

Operations

bgp

read, write

Examples

The following example shows how to redistribute IP Version 4 (IPv4) unicast OSPF routes from OSPF instance 110 into BGP:

Router(config)# router bgp 109
Router(config-bgp)# address-family ipv4 unicast
Router(config-bgp-af)# redistribute ospf 110

refresh-time (rpki-server)

To configure the time BGP waits in between sending periodic serial queries to the RPKI server, use the refresh-time command in rpki-server configuration mode. To remove the refresh-time configuration, use the no form of this command.

refresh-time { time-in-seconds | off }

no refresh-time { time-in-seconds | off }

Syntax Description

off

Specifies not to send serial queries periodically.

time-in-seconds

Sets the refresh-time in seconds. Range is 30 to 3600 seconds.

Command Default

Sets a default value of 600 seconds.

Command Modes

RPKI cache configuration

Command History

Release Modification

Release 4.2.1

This command was introduced.

Usage Guidelines

To use this command, you must be in a user group associated with a task group that includes appropriate task IDs. If the user group assignment is preventing you from using a command, contact your AAA administrator for assistance.

Task ID

Task ID Operation

bgp

read, write

Examples

This example shows how to set the refresh-time for BGP to wait in between sending periodic serial queries to the server as 30 seconds:


Router#configure
Router(config)#router bgp 100
Router(config-bgp)#rpki server 172.168.35.40 
Router(config-bgp-rpki-server)# transport ssh port 22
Router(config-bgp-rpki-server)#username rpki-user
Router(config-bgp-rpki-server)#password rpki-ssh-pass
Router(config-bgp-rpki-server)#preference 1
Router(config-bgp-rpki-server)#purge-time 30
Router(config-bgp-rpki-server)#refresh-time  30

remote-as (BGP)

To create a Border Gateway Protocol (BGP) neighbor and begin the exchange of routing information, use the remote-as command in an appropriate configuration mode. To delete the entry for the BGP neighbor, use the no form of this command.

remote-as as-number

no remote-as [as-number]

Syntax Description

as-number

Autonomous system (AS) to which the neighbor belongs.

  • Range for 2-byte Autonomous system numbers (ASNs) is 1 to 65535.

  • Range for 4-byte Autonomous system numbers (ASNs) in asplain format is 1 to 4294967295.

  • Range for 4-byte Autonomous system numbers (ASNs) is asdot format is 1.0 to 65535.65535.

Command Default

No BGP neighbors exist.

Command Modes

Neighbor configuration

VRF neighbor configuration

Neighbor group configuration

Session group configuration

Command History

Release

Modification

Release 3.7.2

This command was introduced.

Release 3.9.0

Asplain format for 4-byte Autonomous system numbers notation was supported.

Usage Guidelines

To use this command, you must be in a user group associated with a task group that includes appropriate task IDs. If the user group assignment is preventing you from using a command, contact your AAA administrator for assistance.

Use the remote-as command to create a neighbor and assign it a remote autonomous system number. A neighbor must have a remote autonomous system number before any other commands can be configured for it. Removing the remote autonomous system from a neighbor causes the neighbor to be deleted. You cannot remove the autonomous system number if the neighbor has other configuration.


Note


We recommend that you use the no neighbor command rather than the no remote-as command to delete a neighbor.


A neighbor specified with a remote autonomous system number that matches the autonomous system number specified in the router bgp command identifies the neighbor as internal to the local autonomous system. Otherwise, the neighbor is considered external.

Configuration of the remote-as command for a neighbor group or session group using the neighbor-group command or session-group command causes all neighbors using the group to inherit the characteristics configured with the command. Configuring the command directly for the neighbor overrides the value inherited from the group.

In the neighbor configuration submode, configuring use of a session group or neighbor group for which remote-as is configured creates a neighbor and assigns it an autonomous system number if the neighbor has not already been created.


Note


Do not combine remote-as commands and no use neighbor-group commands, or remote-as commands and no use session-group commands, in the same configuration commit.


Task ID

Task ID

Operations

bgp

read, write

Examples

The following example shows how to assign autonomous system numbers on two neighbors, neighbor 10.0.0.1, (internal) and neighbor 192.168.0.1 (external), setting up a peering session that shares routing information between this router and each of these neighbors:


Router(config)# router bgp 1
Router(config-bgp)# session-group group2
Router(config-bgp-sngrp)# remote-as 1
Router(config-bgp-sngrp)#exit 
Router(config-bgp)# neighbor 10.0.0.1
Router(config-bgp-nbr)#use session-group group2 

The following example shows how to configure a session group called group2 with an autonomous system number 1. Neighbor 10.0.0.1 is created when it inherits the autonomous system number 1 from session group group2.


Router(config)#router bgp 1 
Router(config-bgp)# session-group group2
Router(config-bgp-sngrp)# remote-as 1
Router(config-bgp-sngrp)# exit
Router(config-bgp)# neighbor 10.0.0.1
Router(config-bgp-nbr)# use session-group group2

remove-private-as

To remove private autonomous system numbers from autonomous system paths when generating updates to external neighbors, use the remove-private-as command in an appropriate configuration mode. To place the router in the default state in which it does not remove private autonomous system numbers, use the no form of this command.

remove-private-as [inheritance-disable] [entire-aspath]

no remove-private-as [inheritance-disable] [entire-aspath]

Syntax Description

inheritance-disable

(Optional) Permits the feature to be disabled from a neighbor group or address family group instead of being inherited.

entire-aspath

(Optional) Removes the entire private autonomous system numbers from an autonomous system path only if all ASes in the path are private.

Command Default

When this command is not specified in the appropriate configuration mode, private autonomous system numbers are not removed from updates sent to external neighbors.

Command Modes

IPv4 address family group configuration

IPv6 address family group configuration

IPv4 neighbor address family configuration

IPv4 neighbor group address family configuration

IPv6 neighbor group address family configuration

VPNv4 neighbor address family configuration

VRF IPv4 neighbor address family configuration

VPNv4 neighbor group address family configuration

VRF IPv6 neighbor address family configuration

Command History

Release

Modification

Release 3.7.2

This command was introduced.

Release 3.9.0

The disable keyword was replaced with the inheritance-disable keyword.

Release 3.9.2

The entire-aspath keyword was supported.

Usage Guidelines

To use this command, you must be in a user group associated with a task group that includes appropriate task IDs. If the user group assignment is preventing you from using a command, contact your AAA administrator for assistance.

This feature is available for external BGP (eBGP) neighbors only.

When an update is passed to the external neighbor, the system drops any private autonomous system numbers. This happens irrespective of whether the autonomous system numbers are at the beginning or in the middle of the AS_SEQUENCE.

If this command is used in a BGP confederation, the element following the confederation portion of the autonomous system path, if a sequence, is considered the leading sequence.

The private autonomous system values range from 64512 to 65535.

If this command is configured for a neighbor group or address family group, all neighbors using the group inherit the configuration. Values of commands configured specifically for a neighbor override inherited values.

Use the entire-aspath to removes the entire private autonomous system numbers from an autonomous system path only if all ASes in the path are private.

Task ID

Task ID

Operations

bgp

read, write

Examples

The following example shows a configuration that removes the private autonomous system number from the IP Version 4 (IPv4) unicast updates sent to 172.20.1.1:

RP/0/RSP0/CPU0:router(config)# router bgp 140
RP/0/RSP0/CPU0:router(config-bgp)# neighbor 172.20.1.1
RP/0/RSP0/CPU0:router(config-bgp-nbr)# remote-as 1
RP/0/RSP0/CPU0:router(config-bgp-nbr)# address-family ipv4 unicast
RP/0/RSP0/CPU0:router(config-bgp-nbr-af)# remove-private-as

The following example shows how to disable the remove private autonomous system number feature for neighbor 172.20.1.1, preventing this feature from being automatically inherited from address family group group1:

RP/0/RSP0/CPU0:router(config)# router bgp 140
RP/0/RSP0/CPU0:router(config-bgp)# af-group group1 address-family ipv4 unicast
RP/0/RSP0/CPU0:router(config-bgp-afgrp)# remove-private-as
RP/0/RSP0/CPU0:router(config-bgp-afgrp)# exit
RP/0/RSP0/CPU0:router(config-bgp)# neighbor 172.20.1.1
RP/0/RSP0/CPU0:router(config-bgp-nbr# remote-as 1
RP/0/RSP0/CPU0:router(config-bgp-nbr)# address-family ipv4 unicast
RP/0/RSP0/CPU0:router(config-bgp-nbr-af)# use af-group group1
RP/0/RSP0/CPU0:router(config-bgp-nbr-af)# remove-private-as inheritance-disable

replace as-path all

To replace BGP AS path with custom values, use the replace as-path all command in route-policy configuration mode.

replace as-path all { none auto | 'x' } [ n ] [ parameter ]

Syntax Description

none

Replaces AS path with null or empty value.

auto

Replaces AS path based on the route policy. For inbound route policy, replaces AS path with AS path of neighbor from where the prefix is received. For outbound route policy, replaces AS path with the configured local AS path.

'x'

Replaces AS path with specified value, where 'x' is a single AS number or a sequence of AS numbers separated by space.

n

(Optional) Repeats replacing the AS path for specified number of times. The range is from 2 to 64. This option is supported only for the auto and 'x' parameters.

parameter

(Optional) Parameter name used along with repeat option. The parameter name must be preceded with a “$”. You can attach the route policy configured with the parameter to a neighbor and specify the number of times the AS path replacement should be repeated.

Command Default

No default behavior or values

Command Modes

Route-policy configuration

Command History

Release

Modification

Release 7.5.2

This command was introduced.

Usage Guidelines

Use the replace as-path all command to replace the BGP AS path with specified custom values.

When you remove the route policy from a neighbor, the associated replace as-path all configuration is also removed.

Task ID

Task ID

Operations

route-policy

read, write

Examples

The following example shows how to use the replace as-path all command to replace the AS path with auto :


Router(config)#route-policy aspath-auto
Router(config-rpl)# replace as-path all auto
Router(config-rpl)# end-policy
Router(config)# router bgp 65530
Router(config-bgp)# neighbor 111.0.0.1
Router(config-bgp-nbr)# address-family ipv4 unicast
Router(config-bgp-nbr-af)# route-policy aspath-auto out

response-time (rpki-server)

To configure the time BGP waits for a response from the RPKI cache-server after sending a serial or reset query, use the response-time command in rpki-server configuration mode. To remove the response-time configuration, use the no form of this command.

response-time { time-in-seconds | off }

no response-time { time-in-seconds | off }

Syntax Description

off

Specifies to wait indefinitely for a response from the RPKI cache.

time-in-seconds

Specifies the response-time in seconds. Range is 30 to 3600 seconds.

Command Default

Sets a default value of 30 seconds.

Command Modes

RPKI server configuration

Command History

Release Modification

Release 4.2.1

This command was introduced.

Usage Guidelines

To use this command, you must be in a user group associated with a task group that includes appropriate task IDs. If the user group assignment is preventing you from using a command, contact your AAA administrator for assistance.

Task ID

Task ID Operation

bgp

read, write

Examples

This example shows how to set the time for BGP to wait for a response from the RPKI server as 30 seconds, after sending a serial or reset query:


Router#configure
Router(config)#router bgp 100
Router(config-bgp)#rpki server 72.168.35.40 
Router(config-bgp-rpki-server)# transport ssh port 22
Router(config-bgp-rpki-server)#username rpki-user
Router(config-bgp-rpki-server)#password rpki-ssh-pass
Router(config-bgp-rpki-server)#preference 1
Router(config-bgp-rpki-server)#purge-time 30
Router(config-bgp-rpki-server)#refresh-time 30
Router(config-bgp-rpki-server)#response-time 30

retain local-label

To retain the local label until the network is converged, use the retain local-label command in an appropriate address family configuration mode. To disable the retaining of the local label, use the no form of this command.

retain local-label minutes

no retain local-label

Syntax Description

minutes

Local retention time in minutes. The range is 3 to 60 minutes. The default retention time is 5 minutes.

Command Default

minutes : 5

Command Modes

L2VPN address family configuration

VPNv4 address family configuration

VPNv6 address family configuration

Command History

Release

Modification

Release 3.9.0

This command was introduced.

Usage Guidelines

To use this command, you must be in a user group associated with a task group that includes appropriate task IDs. If the user group assignment is preventing you from using a command, contact your AAA administrator for assistance.

Task ID

Task ID

Operations

bgp

read, write

Examples

The following example shows how to enable local label retention for 5 minutes:


RP/0/RSP0/CPU0:routerconfigure
RP/0/RSP0/CPU0:router(config)# router bgp 100
RP/0/RSP0/CPU0:router(config-bgp)# address-family vpnv4 unicast
RP/0/RSP0/CPU0:router(config-bgp-af)# retain local-label 5

retain route-target

To accept received updates with specified route targets, use the retain route-target command in an appropriate configuration mode. To disable the retaining of routes tagged with specified route targets, use the no form of this command.

retain route-target {all | route-policy route-policy-name}

no retain route-target [all | route-policy route-policy-name]

Syntax Description

all

Accepts received updates containing at least one route target.

route-policy router-policy-name

Accepts received updates accepted by a specified route filter policy.

Command Default

The default is to accept all route targets.

Command Modes

VPNv4 address family configuration

Command History

Release

Modification

Release 3.7.2

This command was introduced.

Usage Guidelines

To use this command, you must be in a user group associated with a task group that includes appropriate task IDs. If the user group assignment is preventing you from using a command, contact your AAA administrator for assistance.

Use the retain route-target command to configure a route reflector (RR) to retain routes tagged with specific route targets (RT).

The retain route-target is a required command for Inter-AS option B ASBR. By default, an Inter-AS option B ASBR needs the retain route-target configured to get VPNv4 BGP table from PE routers, either with the all or with the route-policy option.

A provider edge (PE) router is not required to hold all VPNv4 routes. The PE router holds only routes that match the import RT of the VPNs configured on it, but a RR must retain all VPNv4 routes because it may peer with PE routers and different PEs may require different RT-tagged VPNv4 routes. Configuring an RR to hold only routes that have a defined set of RT communities and configuring some of these RRs to service a different set of VPNs provides scalability to the RRs. A PE can be configured to peer with all RRs that service the VPN routing and forwarding (VRF) instances configured on the PE. When a new VRF is configured with an RT for which the PE does not already hold routes, the PE issues route refresh requests to the RRs and gets the relevant VPN routes.

The route-policy route-policy-name keyword and argument takes the policy name that lists the extended communities that a path should have for the RR to retain the path.

Task ID

Task ID

Operations

bgp

read, write

Examples

The following example shows how to configure RR to retain all routes with the route filter policy ft-policy-A:

RP/0/RSP0/CPU0:router(config)# router bgp 140
RP/0/RSP0/CPU0:router(config-bgp)# address-family vpnv4 unicast
RP/0/RSP0/CPU0:router(config-bgp-af)# retain route-target route-filter ft-policy-A

route-monitoring inbound post-policy

To enable the Adj-RIB-In post-policy (inbound post-policy) view for all the address families, run the route-monitoring inbound post-policy command in router configuration mode. Use the no form of the command to disable the feature.

route-monitoring inbound post-policy

Syntax Description

inbound post-policy

Inbound post-policy route-monitoring mode

Command Default

None

Command Modes

Router configuration

Command History

Release

Modification

Release 7.5.4

The syntax of this command was modified to enable the Adj-RIB-In post-policy (inbound post-policy) view for all the address families.

Release 5.2.2

This command was introduced.

Usage Guidelines

  • To use this command, you must be in a user group associated with a task group that includes appropriate task IDs. If the user group assignment is preventing you from using a command, contact your AAA administrator for assistance.

  • Adj-RIB-In Pre-policy view for all the AFIs is available by default if at least one BMP server is configured.

  • Adj-RIB-In Pre-policy cannot be configured with route-monitoring inbound post-policy or Local-RIB.

Task ID

Task ID

Operations

bgp

read, write

Examples

The following example shows how to configure the route-monitoring inbound post-policy:
Router#config 
Router(config)#bmp server all 
Router(config-bgp-bmp)#route-monitoring inbound post-policy  
Router(config-bgp-bmp-rmon)#commit 

route-monitoring local-rib

To enable the local RIB view for all the address families, run the route-monitoring local-rib command in router configuration mode. Use the no form of the command to disable the feature.

route-monitoring local-rib

Syntax Description

local-rib

Local-RIB route-monitoring mode

Command Default

None

Command Modes

Router configuration

Command History

Release

Modification

Release 7.5.4

This command was introduced.

Usage Guidelines

To use this command, you must be in a user group associated with a task group that includes appropriate task IDs. If the user group assignment is preventing you from using a command, contact your AAA administrator for assistance.

Task ID

Task ID

Operations

bgp

read, write

Examples

The following example shows how to configure the route-monitoring local-rib:
Router#config 
Router(config)#bmp server all 
Router(config-bgp-bmp)#route-monitoring local-rib  
Router(config-bgp-bmp-rmon)#commit

route-policy (BGP)

To apply a routing policy to updates advertised to or received from a Border Gateway Protocol (BGP) neighbor, use the route-policy command in an appropriate configuration mode. To disable applying routing policy to updates, use the no form of this command.

route-policy route-policy-name [ parameter1, parameter2, . . . , parametern ] { in | | out }

no route-policy route-policy-name [ parameter1, parameter2, . . . , parametern ] { in | | out }

route-policy aggregate-route-policy-name [ set aggregate-contributor ]

route-policy [set_comm_in | set_comm_out]

Syntax Description

route-policy-name

Name of route policy. Up to 16 parameters can follow the route-policy-name, enclosed in brackets ([ ]).

in

Applies policy to inbound routes.

out

Applies policy to outbound routes.

set_comm_in

Apply the inbound policy to the neighbors after setting the BGP attributes to the aggregate-contributor.

set_comm_out

Apply the outbound policy to the neighbors after setting the BGP attributes to the aggregate-contributor.

set aggregate-contributor

Sets the specific routes as aggregate contributors.

Command Default

No policy is applied.

Command Modes

IPv4 address family group configuration

IPv6 address family group configuration

IPv4 neighbor address family configuration

IPv4 neighbor group address family configuration

IPv6 neighbor group address family configuration

VPNv4 address family group configuration

VPNv4 neighbor address family configuration

VRF IPv4 neighbor address family configuration

VPNv4 neighbor group address family configuration

VRF IPv6 neighbor address family configuration

Command History

Release

Modification

Release 3.7.2

This command was introduced.

Release 7.5.4

The set aggregate-contributor keyword is added.

Usage Guidelines

To use this command, you must be in a user group associated with a task group that includes appropriate task IDs. If the user group assignment is preventing you from using a command, contact your AAA administrator for assistance.

Use the route-policy command to specify a routing policy for an inbound or outbound route. The policy can be used to filter routes or modify route attributes. The route-policy command is used to define a policy.


Note


Configuring a large number of uniquely named outbound neighbor policies can adversely affect performance. This is true even if the uniquely named route policies are functionally identical. The user is discouraged from configuring multiple functionally identical route policies for use with this command. For example, if Policy A and Policy B are identical but named for different neighbors, the two policies should be configured as a single policy.


If the route-policy command is configured for a neighbor group or neighbor address family group, all neighbors using the group inherit the configuration. Values of commands configured specifically for a neighbor override inherited values.

The aggregate-contributor feature is applicable for the following Address Family Indicators (AFIs):

  • Ipv4 unicast

  • IPv6 unicast

Aggregate-contributor flag is set in the router where you configure it.

Task ID

Task ID

Operations

bgp

read, write

Examples

The following example shows how to apply the In-Ipv4 policy to inbound IP Version 4 (IPv4) unicast routes from neighbor 172.20.1.1:

Router(config)# router bgp 1
Router(config-bgp)# neighbor 172.20.1.1
Router(config-bgp-nbr)# remote-as 1
Router(config-bgp-nbr)# address-family ipv4 unicast
Router(config-bgp-nbr-af)# route-policy In-Ipv4 in

The following example shows how to set an aggregate contributor to a route policy:


Router# config
Router(config)#route-policy aggregate-policy1
Router(config-rpl)#set aggregate-contributor
Router(config-rpl)#end
Router(config-rpl)#commit

route-reflector-client

To configure the router as a Border Gateway Protocol (BGP) route reflector and configure the specified neighbor as its client, use the route-reflector-client command in an appropriate configuration mode. To disable configuring the neighbor as a client, use the no form of this command.

route-reflector-client [inheritance-disable]

no route-reflector-client [inheritance-disable]

Syntax Description

inheritance-disable

(Optional) Allows the configuration inherited from a neighbor group or address family group to be overridden.

Command Default

The neighbor is not treated as a route reflector client.

Command Modes

IPv4 address family group configuration

IPv6 address family group configuration

IPv4 neighbor address family configuration

IPv4 neighbor group address family configuration

IPv6 neighbor group address family configuration

VPNv4 address family group configuration

VPNv4 neighbor address family configuration

VPNv4 neighbor group address family configuration

Command History

Release

Modification

Release 3.7.2

This command was introduced.

Release 3.9.0

The disable keyword was replaced with the inheritance-disable keyword.

Usage Guidelines

To use this command, you must be in a user group associated with a task group that includes appropriate task IDs. If the user group assignment is preventing you from using a command, contact your AAA administrator for assistance.

This command is restricted to internal BGP (iBGP) neighbors only.

Use the route-reflector-client command to configure the local router as the route reflector and the specified neighbor as one of its clients. All neighbors configured with this command are members of the client group, and the remaining iBGP peers are members of the nonclient group for the local route reflector.

By default, all iBGP speakers in an autonomous system must be fully meshed with each other, and neighbors do not readvertise iBGP learned routes to other iBGP neighbors.

With route reflection, all iBGP speakers need not be fully meshed. An iBGP speaker, the route reflector, passes learned iBGP routes to some number of iBGP client neighbors. Learned iBGP routes eliminate the need for each router running BGP to communicate with every other device running BGP in the autonomous system.

The local router is a route reflector as long as it has at least one route reflector client.

If this command is configured for a neighbor group or neighbor address family group, all neighbors using the group inherit the configuration. Values of commands configured specifically for a neighbor override inherited values.

Task ID

Task ID

Operations

bgp

read, write

Examples

The following example shows neighbor at 172.20.1.1 configured as a route reflector client for IP Version 4 (IPv4) unicast routes:

RP/0/RSP0/CPU0:router(config)# router bgp 140
RP/0/RSP0/CPU0:router(config-bgp)# neighbor 172.20.1.1
RP/0/RSP0/CPU0:router(config-bgp-nbr)# remote-as 140
RP/0/RSP0/CPU0:router(config-bgp-nbr)# address-family ipv4 unicast
RP/0/RSP0/CPU0:router(config-bgp-nbr-af)# route-reflector-client

The following example disables the route-reflector client for neighbor 172.20.1.1, preventing this feature from being automatically inherited from address family group group1:

RP/0/RSP0/CPU0:router(config)# router bgp 140
RP/0/RSP0/CPU0:router(config-bgp)# af-group group1 address-family ipv4 unicast
RP/0/RSP0/CPU0:router(config-bgp-afgrp)# route-reflector-client
RP/0/RSP0/CPU0:router(config-bgp-afgrp)#exit
RP/0/RSP0/CPU0:router(config-bgp)# neighbor 172.20.1.1
RP/0/RSP0/CPU0:router(config-bgp-nbr)# remote-as 140
RP/0/RSP0/CPU0:router(config-bgp-nbr)# address-family ipv4 unicast
RP/0/RSP0/CPU0:router(config-bgp-nbr-af)# use af-group group1
RP/0/RSP0/CPU0:router(config-bgp-nbr-af)# route-reflector-client inheritance-disable

optimal-route-reflection

To enable the BGP optimal route reflector (ORR) feature, use the optimal-route-reflection command in router BGP, or neighbor configuration mode, as appropriate.

optimal-route-reflection orr-group-name primary-ip-address [secondary-ip-address] [tertiary-ip-address] router-policy { addpath name }

Syntax Description

orr-group-name

Specify the ORR group name. A maximum of 32 characters are allowed.

primary-ip-address

Specify the primary SPF root IP address. Depending on the address family configured under BGP, the SPF root IP address can be either IPv4 or IPv6.

The primary SPF root IP address is the IP address of the router for which best path is calculated.

secondary-ip-address

[Optional] Specify the secondary SPF root IP address. Depending on the address family configured under BGP, the SPF root IP address can be either IPv4 or IPv6.

The secondary SPF root IP address is the IP address of the nearest neighbor of the router for which best path is calculated.

tertiary-ip-address

[Optional] Specify the tertiary SPF root IP address. Depending on the address family configured under BGP, the SPF root IP address can be either IPv4 or IPv6.

The secondary SPF root IP address is the IP address of the nearest neighbor of the router for which best path is calculated.

router-policy

[Optional] Enables the ORR group add path policy.

addpath name

[Optional] Specify the ORR group add path policy name.

Command Default

BGP ORR is disabled by default.

Command Modes

router BGP

neighbor configuration

Command History

Release Modification

Release 24.2.1

The keyword router-policy was added.

Release 6.1.2

This command was introduced.

Usage Guidelines

Task ID

Task ID Operation

bgp

read, write

Examples

This sample shows how to determine shortest exit point for the router with IP address 192.0.2.1, in the domain with AS number 6500, and ORR group name group1. This configuration is executed on virtual router reflector:


vRR# router bgp 6500
      address-family ipv4 unicast
        optimal-route-reflection group1 192.0.2.1
     commit

Examples

The following example shows how to configure the an ORR group add path policy:
router# configure
router(config)# router bgp 13
router(config-bgp)# optimal-route-reflection ipv4 orr-grp-1 1.1.1.1 route-policy addpath-backup
router(config-bgp-router)# address-family ipv4 unicast
router(config-bgp-router-af)# additional-paths selection route-policy addpath-all
router(config-bgp-router-af)# optimal-route-reflection apply orr-grp-1
router(config-bgp-router-af)# commit

router bgp

To configure the Border Gateway Protocol (BGP) routing process, use the router bgp command in Global Configuration mode mode. To remove all BGP configurations and terminate the BGP routing process, use the no form of this command.

router bgp as-number

Syntax Description

as-number

Number that identifies the autonomous system (AS) in which the router resides.

  • Range for 2-byte Autonomous system numbers (ASNs) is 1 to 65535.

  • Range for 4-byte Autonomous system numbers (ASNs) in asplain format is 1 to 4294967295.

  • Range for 4-byte Autonomous system numbers (ASNs) is asdot format is 1.0 to 65535.65535.

Command Default

No BGP routing process is enabled.

Command Modes

Global Configuration mode

Command History

Release

Modification

Release 3.7.2

This command was introduced.

Release 3.9.0

Asplain format for 4-byte Autonomous system number notation was supported.

Usage Guidelines

Use the router bgp command to set up a distributed routing core that automatically guarantees the loop-free exchange of routing information between autonomous systems.

Task ID

Task ID

Operations

bgp

read, write

rib

read, write

Examples

The following example shows how to configure a BGP process for autonomous system 120:

RP/0/RSP0/CPU0:router(config)# router bgp 120

rpki server

To enter resource public key infrastructure (RPKI) cache-server (rpki-sever) configuration mode and enable rpki parameters configuration, use the rpki server command in Router BGP configuration mode. To remove the rpki-server configuration mode and delink cache-server from the cache list, use the no form of this command.

rpki server { host-name | ip-address }

no rpki server { host-name | ip-address }

Syntax Description

host-name

Host name of the RPKI cache database.

ip-address

IP Address of the RPKI cache databse.

Command Default

RPKI server configuration is disabled.

Command Modes

Router BGP configuration

Command History

Release Modification

Release 4.2.1

This command was introduced.

Usage Guidelines

To use this command, you must be in a user group associated with a task group that includes appropriate task IDs. If the user group assignment is preventing you from using a command, contact your AAA administrator for assistance.

Task ID

Task ID Operation

bgp

read, write

Examples

This example shows how to configure an rpki cache-server database and enter rpki-server configuration mode:


Router#configure
Router(config)#router bgp 100
Router(config-bgp)#rpki server 172.168.35.40
Router(config-bgp-rpki-cache)#

rpki route

To statically configure an RPKI route, use the rpki route command in the router BGP configuration submode. The no form of this command removes the RPKI routes.

rpki route ip-address-length {max max-prefix-length | origin origin-autonomous-system-number}

no rpki route ip-address-length {max max-prefix-length | origin origin-autonomous-system-number}

Syntax Description

ip-address/length

Specifies the IP address of the network along with the minimum prefix length.

max max-prefix-length Specifies the maximum prefix length (32 for IPv4 and 128 for IPv6).
origin origin-autonomous-system-number Specifies the autonomous system number.

Command Default

RPKI route configuration is disabled.

Command Modes

Router BGP configuration

Command History

Release Modification
Release 4.2.1

This command was introduced.

Usage Guidelines

To use this command, you must be in a user group associated with a task group that includes appropriate task IDs. If the user group assignment is preventing you from using a command, contact your AAA administrator for assistance.

In general, BGP receives the Route-Origin-Attestation (ROA) information from RPKI cache. However, the rpki route command is used for verification. This command can be used to configure both IPv4 and IPv6 ROAs.

This command contains all the essential attributes of an ROA record, that is, the prefix-block (IP address/length (minimum/maximum)) and the origin AS authorized to create the prefix-block.

Multiple static ROAs can be configured through this command and these entries will be included in the routers RPKI database, as if they were fetched from an RPKI cache.

Task ID

Task ID Operation

bgp

read, write

Examples

This example shows how to configure an rpki route:


RP/0/RSP0/CPU0:router#configure
RP/0/RSP0/CPU0:router(config)#router bgp 100
RP/0/RSP0/CPU0:router(config-bgp)#rpki route 192.168.1.0/24 max 30 origin 65001
RP/0/RSP0/CPU0:router(config-bgp)#rpki route 172.200.0.0/16 max 24 origin 300
RP/0/RSP0/CPU0:router(config-bgp)#

selective-vrf-download disable

To disable selective VRF download (SVD) on a line card to enable download all prefixes and labels to the line card, use the selective-vrf-download disable command in global configuration mode. To enable the SVD, use the no form of this command.

selective-vrf-download disable

no selective-vrf-download disable

Syntax Description

This command has no keywords or arguments.

Command Default

SVD is enabled.

Command Modes

global configuration

Command History

Release Modification

Release 4.2.0

This command was introduced.

Usage Guidelines

To use this command, you must be in a user group associated with a task group that includes appropriate task IDs. If the user group assignment is preventing you from using a command, contact your AAA administrator for assistance.

You must failover the active RP or reload the router after disabling SVD for the configuration change to get activated.

Task ID

Task ID Operation

ip-services

read, write

Examples

This example shows how to disable selective vrf download:


RP/0/RSP0/CPU0:router#configure
RP/0/RSP0/CPU0:router(config)#selective-vrf-download disable 

send-buffer-size

To set the size of the send buffers for a Border Gateway Protocol (BGP) neighbor, use the send-buffer-size command in an appropriate configuration mode. To set the size of the send buffers to the default values, use the no form of this command.

send-buffer-size socket-size [bgp-size]

no send-buffer-size [socket-size] [bgp-size]

Syntax Description

socket-size

Size, in bytes, of the send-side socket buffer. Range is 4096 to 131072.

bgp-size

(Optional) Size, in bytes, of the BGP process send buffer. Range is 4096 to 131072.

Command Default

socket-size : 10240 bytes

bgp-size : 4096 bytes

Use the socket send-buffer-size command to change the defaults.

Command Modes

Neighbor configuration

VRF neighbor configuration

Neighbor group configuration

Session group configuration

Command History

Release

Modification

Release 3.7.2

This command was introduced.

Usage Guidelines

To use this command, you must be in a user group associated with a task group that includes appropriate task IDs. If the user group assignment is preventing you from using a command, contact your AAA administrator for assistance.

Use the send-buffer-size command to increase the buffer size employed when sending updates to a neighbor. Using larger buffers can improve convergence time because the software can process more packets simultaneously. However, allocating larger buffers uses more memory on the router.


Note


Increasing the socket buffer size uses more memory only when more messages are waiting to be processed by the software. In contrast, increasing the BGP buffer size uses more memory indefinitely.


If this command is configured for a neighbor group or session group, all neighbors using the group inherit the configuration. Values of commands configured specifically for a neighbor override inherited values.

Task ID

Task ID

Operations

bgp

read, write

Examples

The following example shows how to set the send buffer sizes for neighbor 172.20.1.1 to be 8192 bytes for both the socket buffer and the BGP buffer:

RP/0/RSP0/CPU0:router(config)# router bgp 1
RP/0/RSP0/CPU0:router(config-bgp)# neighbor 172.20.1.1
RP/0/RSP0/CPU0:router(config-bgp-nbr)# remote-as 1
RP/0/RSP0/CPU0:router(config-bgp-nbr)# send-buffer-size 8192 8192

send-community-ebgp

To specify that community or the large community attributes should be sent to an external Border Gateway Protocol (eBGP) neighbor, use the send-community-ebgp command in an appropriate configuration mode. To disable sending community or large community attributes to an eBGP neighbor, use the no form of this command.

send-community-ebgp [inheritance-disable]

no send-community-ebgp [inheritance-disable]

Syntax Description

inheritance-disable

(Optional) Allows configuration inherited from a neighbor group or address family group to be overridden.

Command Default

Community (COMM) attributes are NOT sent to eBGP peers (including PE-CE peers).

Command Modes

IPv4 address family group configuration

IPv6 address family group configuration

IPv4 neighbor address family configuration

IPv4 neighbor group address family configuration

IPv6 neighbor group address family configuration

VRF IPv4 neighbor address family configuration

VPNv4 neighbor address family configuration

VRF IPv6 neighbor address family configuration

VPNv6 neighbor address family configuration

Command History

Release

Modification

Release 3.7.2

This command was introduced.

Release 3.9.0

The disable keyword was replaced with the inheritance-disable keyword.

Release 6.3.1

This command supports large communities.

Usage Guidelines

To use this command, you must be in a user group associated with a task group that includes appropriate task IDs. If the user group assignment is preventing you from using a command, contact your AAA administrator for assistance.

Use the send-community-ebgp command to control whether community or large community attributes are sent to eBGP neighbors. This command cannot be configured for iBGP neighbors as community or large community attributes are always sent to iBGP neighbors.

When IOS XR BGP updates community or large community attributes for eBGP VPN peers (VPNv4 or VPNv6), there is no need to configure the send-community-ebgp command separately. The community attributes are updated by default.

If this command is configured for a neighbor group or address family group, all neighbors using the group inherit the configuration. Configuring the command specifically for a neighbor overrides inherited values.

Task ID

Task ID

Operations

bgp

read, write

Examples

The following example shows how to disable the router that sends community attributes to neighbor 172.20.1.1 for IP Version 4 (IPv4) multicast routes:


RP/0/RSP0/CPU0:router(config)#router bgp 140 
RP/0/RSP0/CPU0:router(config-bgp)# neighbor 172.20.1.1
RP/0/RSP0/CPU0:router(config-bgp-nbr)# remote-as 1
RP/0/RSP0/CPU0:router(config-bgp-nbr)# address-family ipv4 multicast
RP/0/RSP0/CPU0:router(config-bgp-nbr-af)# send-community-ebgp

The following example shows how to disable the delivery of community attributes to neighbor 172.20.1.1, preventing this feature from being inherited from address family group group1:


RP/0/RSP0/CPU0:router(config)#router bgp 140 
RP/0/RSP0/CPU0:router(config-bgp)# af-group group1 address-family ipv4 multicast
RP/0/RSP0/CPU0:router(config-bgp-afgrp)# send-community-ebgp
RP/0/RSP0/CPU0:router(config-bgp-afgrp)# exit
RP/0/RSP0/CPU0:router(config-bgp)# neighbor 172.20.1.1
RP/0/RSP0/CPU0:router(config-bgp-nbr)# remote-as 1
RP/0/RSP0/CPU0:router(config-bgp-nbr)# address-family ipv4 multicast
RP/0/RSP0/CPU0:router(config-bgp-nbr-af)# use af-group group1
RP/0/RSP0/CPU0:router(config-bgp-nbr-af)# send-community-ebgp inheritance-disable

send-community-gshut-ebgp

To direct the router to add the gshut community to the path having the gshut attribute or the path being sent to a connection that has graceful maintenance activated, use the send-community-gshut-ebgp command in the neighbor address family configuration mode. To disable the g-shut community from being announced to ebgp neighbors, use the no form of this command.

send-community-gshut-ebgp [inheritance-disable]

Syntax Description

inheritance-disable

(Optional) Prevent send-community-gshut-ebgp from being inherited from the parent.

Command Default

g-shut community attribute is not sent to eBGP neighbors.

Command Modes

IPv4 address family group configuration

IPv6 address family group configuration

IPv4 neighbor address family configuration

IPv4 neighbor group address family configuration

IPv6 neighbor group address family configuration

Command History

Release

Modification

Release 5.3.2

This command was introduced.

Usage Guidelines

Under neighbor address family configuration, use the send-community-gshut-ebgp command to allow the g-shut community to be sent if it is an ebgp neighbor. A path acquires the gshut attribute when it is received from a connection that has graceful maintenance activated. The sending of the gshut community if it is present because the path was received with that community or if it was added by outbound policy is governed like all other communities by the send-community-ebgp configuration.

Task ID

Task ID

Operations

bgp

read, write

send-extended-community-ebgp

To specify that extended community attributes should be sent to external Border Gateway Protocol (eBGP) neighbors, use the send-extended-community-ebgp command in an appropriate configuration mode. To disable sending extended community attributes to eBGP neighbors, use the no form of this command.

send-extended-community-ebgp [inheritance-disable]

no send-extended-community-ebgp [inheritance-disable]

Syntax Description

inheritance-disable

(Optional) Allows configurations inherited from a neighbor group or address family group to be overridden.

Command Default

Extended community (EXTCOMM) attributes are NOT sent to eBGP peers (including PE-CE peers).

Command Modes

IPv4 address family group configuration

IPv6 address family group configuration

IPv4 neighbor address family configuration

IPv4 neighbor group address family configuration

IPv6 neighbor group address family configuration

VRF IPv4 neighbor address family configuration

VPNv4 neighbor address family configuration

VRF IPv6 neighbor address family configuration

VPNv6 neighbor address family configuration

Command History

Release

Modification

Release 3.7.2

This command was introduced.

Release 3.9.0

The disable keyword was replaced with the inheritance-disable keyword.

Usage Guidelines

To use this command, you must be in a user group associated with a task group that includes appropriate task IDs. If the user group assignment is preventing you from using a command, contact your AAA administrator for assistance.

Use the send-extended-community-ebgp command to control whether extended community attributes are sent to eBGP neighbors. This command cannot be used for iBGP neighbors as extended community attributes are always sent to iBGP neighbors.

When IOS XR BGP updates community attributes for eBGP VPN peers (VPNv4 or VPNv6), there is no need to configure the send-extended-community-ebgp command separately. The community attributes are updated by default.

If this command is configured for a neighbor group or neighbor address family group, all neighbors using the group inherit the configuration. Values of commands configured specifically for a neighbor override inherited values.

Task ID

Task ID

Operations

bgp

read, write

Examples

The following example shows how to configure the router to send extended community attributes to neighbor 172.20.1.1 for IP Version 4 (IPv4) multicast routes:

RP/0/RSP0/CPU0:router(config)# router bgp 140
RP/0/RSP0/CPU0:router(config-bgp)# neighbor 172.20.1.1
RP/0/RSP0/CPU0:router(config-bgp-nbr)# remote-as 1
RP/0/RSP0/CPU0:router(config-bgp-nbr)# address-family ipv4 multicast
RP/0/RSP0/CPU0:router(config-bgp-nbr-af)# send-extended-community-ebgp

The following example shows how to disable the delivery of extended community attributes to neighbor 172.20.1.1, preventing this feature from being automatically inherited from address family group group1:

RP/0/RSP0/CPU0:router(config)# router bgp 140
RP/0/RSP0/CPU0:router(config-bgp)# af-group group1 address-family ipv4 multicast
RP/0/RSP0/CPU0:router(config-bgp-afgrp)# send-extended-community-ebgp
RP/0/RSP0/CPU0:router(config-bgp-afgrp)# exit
RP/0/RSP0/CPU0:router(config-bgp)# neighbor 172.20.1.1
RP/0/RSP0/CPU0:router(config-bgp-nbr)# remote-as 1
RP/0/RSP0/CPU0:router(config-bgp-nbr)# address-family ipv4 multicast
RP/0/RSP0/CPU0:router(config-bgp-nbr-af)# use af-group group1
RP/0/RSP0/CPU0:router(config-bgp-nbr-af)# send-extended-community-ebgp inheritance-disable

session-group

To create a session group and enter session group configuration mode, use the session-group command in router configuration mode. To remove a session group and delete all configurations associated with it, use the no form of this command.

session-group name

no session-group name

Syntax Description

name

Name of the session group.

Command Default

No session groups are created.

Command Modes

Router configuration

Command History

Release

Modification

Release 3.7.2

This command was introduced.

Usage Guidelines

To use this command, you must be in a user group associated with a task group that includes appropriate task IDs. If the user group assignment is preventing you from using a command, contact your AAA administrator for assistance.

Use the session-group command to create a session group from which neighbors can inherit configuration that is address family-independent. That is, session groups cannot have address family-specific configuration. This command enters the session group configuration mode in which configuration for a session group is entered.

Many commands can be configured in both session group configuration mode and neighbor configuration mode.

Use of session groups saves time and reduces the router configuration size. Because the configuration of a session group can be inherited by any number of neighbors, use of the group can eliminate the need to copy long or complex configurations on each of a large number of neighbors. A neighbor can inherit all configuration from a session group simply by configuring the use command. Specific inherited session group configuration commands can be overridden for a specific neighbor by explicitly configuring the command for the specific neighbor.

The no form of this command causes all of the configuration for the session group to be removed. You cannot use the no form of this command if removing the group would leave one or more neighbors without a configured remote autonomous system number.

Task ID

Task ID

Operations

bgp

read, write

Examples

The following example shows a session group called group1 that is used by two neighbors, 10.0.0.1 and 10.0.0.2. Because group1 is a session group, it contains only address family-independent configuration. And because group1 is used by neighbors 10.0.0.1 and 10.0.0.2, they inherit the configuration of the group.

RP/0/RSP0RP0/CPU0:router(config)# router bgp 1
RP/0/RSP0RP0/CPU0:router(config-bgp)# session-group group1
RP/0/RSP0RP0/CPU0:router(config-bgp-sngrp)# remote-as 1
RP/0/RSP0RP0/CPU0:router(config-bgp-sngrp)# advertisement-interval 2
RP/0/RSP0RP0/CPU0:router(config-bgp-sngrp)# exit
RP/0/RSP0RP0/CPU0:router(config-bgp)# neighbor 10.0.0.1
RP/0/RSP0RP0/CPU0:router(config-bgp-nbr)# use session-group group1
RP/0/RSP0RP0/CPU0:router(config-bgp-nbr)# exit
RP/0/RSP0RP0/CPU0:router(config-bgp)# neighbor 10.0.0.2
RP/0/RSP0RP0/CPU0:router(config-bgp-nbr)# use session-group group1

The following example shows a session group called group1 used by two neighbors, 10.0.0.1 and 10.0.0.2. Because group1 is a session group, it contains only address family-independent configuration. And because group1 is used by neighbors 10.0.0.1 and 10.0.0.2, they inherit the configuration of the group. However, the password password1 configuration from group1 is overridden for neighbor 10.0.0.2, using the password-disable command in the neighbor 10.0.0.2 configuration submode.

Router(config)# router bgp 1
Router(config-bgp)# session-group group1
Router(config-bgp-sngrp)# remote-as 1
Router(config-bgp-sngrp)# advertisement-interval 2
Router(config-bgp-sngrp)# password password1
Router(config-bgp-sngrp)# exit
Router(config-bgp)# neighbor 10.0.0.1
Router(config-bgp-nbr)# use session-group group1
Router(config-bgp-nbr)# exit
Router(config-bgp)# neighbor 10.0.0.2
Router(config-bgp-nbr)# use session-group group1
Router(config-bgp-nbr)# password-disable

session-open-mode

To establish a Border Gateway Protocol (BGP) session with a specific TCP open mode, use the session-open-mode command in an appropriate configuration mode. To restore the default state, use the no form of this command.

session-open-mode {active-only | both | passive-only}

no session-open-mode [active-only | both | passive-only]

Syntax Description

active-only

Ensures that the BGP session can be established only when the request is initiated by the local end (active-open request) and all passive-open requests (from the other end) are rejected by the local BGP.

both

Allows BGP sessions to be established from both incoming or outgoing TCP connection requests, with one being rejected in the event of a request collision.

passive-only

Ensures that the local BGP does not initiate any TCP open requests and the session can be established only when the request comes from the remote end.

Command Default

The default is both .

Command Modes

Neighbor configuration

VRF neighbor configuration

Neighbor group configuration

Session group configuration

Command History

Release

Modification

Release 3.7.2

This command was introduced.

Usage Guidelines

To use this command, you must be in a user group associated with a task group that includes appropriate task IDs. If the user group assignment is preventing you from using a command, contact your AAA administrator for assistance.

BGP, by default, tries to initiate an active TCP connection whenever a new neighbor is configured. A remote neighbor may also initiate the TCP connection before the local BGP can initiate the connection. This initiation of a TCP connection by a remote neighbor is considered a passive-open request and it is accepted by the local BGP. This default behavior can be modified using the session-open-mode command.


Note


The BGP connection is not opened and, as a result the BGP session, is not established if both the peering neighbors use the same nondefault TCP session open mode—active-only or passive-only. If both ends are configured with active-only, each neighbor rejects the TCP open request from the other end. One neighbor must be configured as passive-only or both. Similarly, if both neighbors are configured with passive-only, neither neighbor initiates the TCP open request and the BGP session is not established. Again, one neighbor must be configured as active-only or both. There is one exception. A connection open request from a neighbor that is configured with the TCP session open mode to be passive-only is processed to detect whether there is a connection collision before the request is rejected. This exception enables the local BGP to reset the session if the remote neighbor goes down and it is not detected by the local router.


Use the session-open-mode command when it may be necessary to preconfigure a neighbor that does not exist. Ensure that BGP does not spend any time actively trying to set up a TCP session with the neighbor. A BGP session does not come up between two neighbors, both of which configure the same nondefault value ( active-only or passive-only keyword) for this command.

Task ID

Task ID

Operations

bgp

read, write

Examples

The following example shows how to enable a BGP session on router bgp 1:

RP/0/RSP0/CPU0:router(config)# router bgp 1
RP/0/RSP0/CPU0:router(config-bgp)# neighbor 45.67.89.01
RP/0/RSP0/CPU0:router(config-bgp-nbr)# session-open-mode active-only

set flow-tag

To set the flow-tag value for the PBR BGP, use the set flow-tag command in route-policy configuration mode.

set flow-tag {number | parameter}

Syntax Description

number

Flow-tag value. Range is from 1 to 63.

parameter

Parameter name. The parameter name must be preceded with a “$.”

Command Default

No default behavior or values

Command Modes

Route-policy configuration

Command History

Release

Modification

Release 5.2.2

This command was introduced.

Usage Guidelines

To use this command, you must be in a user group associated with a task group that includes appropriate task IDs. If the user group assignment is preventing you from using a command, contact your AAA administrator for assistance.

Use the set flow-tag command to set the flow-tag to classify packets.

This command is supported at the BGP table-policy attachpoint. Prefixes are marked for subsequent processing in the forwarding plane. After flow-tag propagation through Border Gateway Protocol (BGP), flow-tag is enabled on an interface, corresponding traffic shaping and policing is completed using packet classification based on the flow-tag value.

Task ID

Task ID

Operations

route-policy

read, write

Examples

This example shows how to use set flow-tag command:


RP/0/RSP0/CPU0:router(config)# route-policy policy_1
RP/0/RSP0/CPU0:router(config-rpl)# set flow-tag 12
RP/0/RSP0/CPU0:router(config-rpl)# end-policy

show bgp

To display entries in the Border Gateway Protocol (BGP) routing table, use the show bgp command in EXEC mode.

show bgp [ipv4 {unicast | multicast | labeled-unicast | all | tunnel} | ipv6 {unicast | multicast | all | labeled-unicast} | all {unicast | multicast | all | labeled-unicast | mdt | tunnel} | vpnv4 { unicast} [rd rd-address] | vrf {vrf-name | all} [ipv4 {unicast | labeled-unicast} | ipv6 {unicast | }] | { unicast} | | [instance] | | [instances] | ] [ip-address [ {mask | / prefix-length} [longer-prefixes | unknown-attributes | bestpath-compare]]] [standby] [detail]

Syntax Description

ipv4

(Optional) Specifies IP Version 4 address prefixes.

unicast

(Optional) Specifies unicast address prefixes.

multicast

(Optional) Specifies multicast address prefixes.

labeled-unicast

(Optional) Specifies labeled unicast address prefixes.

all

(Optional) For subaddress families, specifies prefixes for all subaddress families.

tunnel

(Optional) Specifies tunnel address prefixes.

ipv6

(Optional) Specifies IP Version 6 address prefixes.

all

(Optional) For address family, specifies prefixes for all address families.

vpnv4 unicast

(Optional) Specifies VPNv4 unicast address families.

instances

(Optional) Displays information of all BGP instances.

vrf

(Optional) Specifies VPN routing and forwarding (VRF) instance.

vrf

(Optional) Specifies VPN routing and forwarding (VRF) instance.

vrf-name

(Optional) Name of a VRF.

all

(Optional) For VRF, specifies all VRFs.

ipv4 { unicast | labeled-unicast }

(Optional) For VRF, specifies IPv4 unicast or labeled-unicast address families.

ipv6 unicast

(Optional) For VRF, specifies IPv6 unicast address families.

ip-address

(Optional) Network address, entered to display a particular network in the BGP routing table. If the network address is omitted, then all networks in the BGP routing table are displayed. If the network mask and prefix length is omitted, then the software displays the longest matching prefix for the network address.

mask

(Optional) Network mask of the BGP route to match.

/ prefix-length

(Optional) Prefix length of the BGP route to match. A slash (/) must precede the decimal value.

longer-prefixes

(Optional) Displays a route with the specified prefix length and more-specific routes if available. The longer-prefixes keyword is available when the ip-address and mask or /prefix-length arguments are specified.

unknown-attributes

(Optional) Includes unknown, transitive attributes. The unknown-attributes keyword is available when the ip-address and mask or /prefix-length arguments are specified.

bestpath-compare

(Optional) Displays route and best-path comparison information. The bestpath-compare keyword is available when the ip-address and mask or /prefix-length arguments are specified.

Command Default

If no address family or subaddress family is specified, the default address family and subaddress family specified using the set default-afi and set default-safi commands are used.

Command Modes

EXEC mode

Command History

Release

Modification

Release 3.7.2

This command was introduced.

Release 3.9.0

Asplain format for 4-byte Autonomous system numbers notation was supported. The input parameters and output were modified to display 4-byte autonomous system numbers and extended communities in either asplain or asdot notations.

Release 5.3.2

The command output was modified to include graceful maintenance feature information.

Release 6.3.1

The command output was modified to include large community in the list of attributes.

Release 7.9.1

The command output was modified to include the aggregate-contributor feature examples.

Usage Guidelines

set default-afi


Note


The set default-afi command is used to specify the default address family for the sessions and the set default-safi command is used to specify the default subaddress family for the session. See the System Management Command Reference for Cisco ASR 9000 Series Routers for detailed information and syntax for the set default-afi and set default-safi commands. If you do not specify a default address family, the default address family is IPv4. If you do not specify a default subaddress family, the default subaddress family is unicast.


BGP contains a separate routing table for each address family and subaddress family combination that has been configured. The address family and subaddress family options specify the routing table to be examined. If the all keyword is specified for an address family or a subaddress family, each matching routing table is examined in turn.


Note


Running the show bgp command immediately after configuring a large and complex route policy may result in timeout of the system database shown through an error message (SYSDB-SYSDB-6-TIMEOUT_EDM). It is recommended, that the show command be run, after the new route policy takes effect.


Use the show bgp ip-address { mask | / prefix-length } command to display detailed information for a specific route. If the mask and prefix length are omitted, the details of the longest matching prefix for the IP address are displayed.

Use the show bgp command to display all routes in the specified BGP routing table. Use the show bgp ip-address { mask | / prefix-length } longer-prefixes command to display those routes more specific than a particular prefix.

Use the unknown-attributes keyword to display details of any transitive attributes associated with a route that are not understood by the local system.

Task ID

Task ID

Operations

bgp

read

Examples

The following is the sample output from the show bgp <IP address> command displaying the graceful-shutdown community and the graceful-shut path attribute with BGP graceful maintenance feature activated:

RP/0/0/CPU0:R4#show bgp 5.5.5.5
...
    10.10.10.1 from 10.10.10.1 (192.168.0.5)
      Received Label 24000
      Origin incomplete, metric 0, localpref 100, valid, internal, best, group-best, import-candidate
      Received Path ID 0, Local Path ID 1, version 4
      Community: graceful-shutdown
      Originator: 192.168.0.5, Cluster list: 192.168.0.1
...

The following is sample output from the show bgp command in EXEC mode:

 
RP/0/RSP0/CPU0:router#show bgp
 BGP router identifier 172.20.1.1, local AS number 1820
 BGP generic scan interval 60 secs
 BGP table state: Active
 Table ID: 0xe0000000
 BGP main routing table version 3
 Dampening enabled
 BGP scan interval 60 secs
 
 Status codes: s suppressed, d damped, h history, * valid, > best
               i - internal, S stale
 Origin codes: i - IGP, e - EGP, ? - incomplete
 
 Network             Next Hop           Metric LocPrf Weight Path
 * i10.3.0.0/16      172.20.22.1             0    100      0 1800 1239 ?
 *>i                 172.20.16.1             0    100      0 1800 1239 ?
 * i10.6.0.0/16      172.20.22.1             0    100      0 1800 690 568 ?
 *>i                 172.20.16.1             0    100      0 1800 690 568 ?
 * i10.7.0.0/16      172.20.22.1             0    100      0 1800 701 35 ?
 *>i                 172.20.16.1             0    100      0 1800 701 35 ?
 *                   192.168.40.24                         0 1878 704 701 35 ?
 * i10.8.0.0/16      172.20.22.1             0    100      0 1800 690 560 ?
 *>i                 172.20.16.1             0    100      0 1800 690 560 ?
 *                   192.168.40.24                         0 1878 704 701 560 ?
 * i10.13.0.0/16     172.20.22.1             0    100      0 1800 690 200 ?
 *>i                 172.20.16.1             0    100      0 1800 690 200 ?
 *                   192.168.40.24                         0 1878 704 701 200 ?
 * i10.15.0.0/16     172.20.22.1             0    100      0 1800 174 ?
 *>i                 172.20.16.1             0    100      0 1800 174 ?
 * i10.16.0.0/16     172.20.22.1             0    100      0 1800 701 i
 *>i                 172.20.16.1             0    100      0 1800 701 i
 *                   192.168.40.24                         0 1878 704 701 i
 
 Processed 8 prefixes, 8 paths
 

This table describes the significant fields shown in the display.

Table 3. show bgp Field Descriptions

Field

Description

BGP router identifier

BGP identifier for the local system.

local AS number

Autonomous system number for the local system.

  • Range for 2-byte Autonomous system numbers (ASNs) is 1 to 65535.
  • Range for 4-byte Autonomous system numbers (ASNs) in asplain format is 1 to 4294967295.
  • Range for 4-byte Autonomous system numbers (ASNs) in asdot format is 1.0 to 65535.65535.

BGP generic scan interval

Interval (in seconds) between scans of the BGP table by a generic scanner.

BGP table state

State of the BGP database.

Table ID

BGP database identifier.

BGP main routing table version

Last version of the BGP database that was installed into the main routing table.

Dampening enabled

Dampening is enabled for the routes in this BGP routing table.

BGP scan interval

Interval (in seconds) between BGP scans for the specified address family and subaddress family.

Status codes

Status of the table entry. The status is displayed as a three-character field at the beginning of each line in the table. The first character may be (in order of precedence):

S—Path is stale, indicating that a graceful restart is in progress with the peer from which the route was learned.

s—Path is more specific than a locally sourced aggregate route and has been suppressed.

*—Path is valid.

The second character may be (in order of precedence):

>—Path is the best path to use for that network.

d—Path is dampened.

h—Path is a history entry, representing a route that is currently withdrawn, but that is being maintained to preserve dampening information. Such routes should never be marked as valid.

The third character may be:

i—Path was learned by an internal BGP (iBGP) session.

Origin codes

Origin of the path. The origin code is displayed at the end of each line in the table. It can be one of the following values:

i—Path originated from an Interior Gateway Protocol (IGP) and was sourced by BGP using a network or aggregate-address command.

e—Path originated from an Exterior Gateway Protocol (EGP).

?—Origin of the path is not clear. Usually, this is a route that is redistributed into BGP from an IGP.

Network

IP prefix and prefix length for a network.

Next Hop

IP address of the next system that is used when a packet is forwarded to the destination network. An entry of 0.0.0.0 indicates that the router has a non-BGP route to this network.

Metric

Value of the interautonomous system metric, otherwise known as the Multi Exit discriminator (MED) metric.

LocPrf

Local preference value. This is used to determine the preferred exit point from the local autonomous system. It is propagated throughout the local autonomous system.

Weight

Path weight. Weight is used in choosing the preferred path to a route. It is not advertised to any neighbor.

Path

Autonomous system path to the destination network. At the end of the path is the path origin code.

The following is sample output from the show bgp command with the network specified:


RP/0/RSP0/CPU0:router# show bgp 11.0.0.0/24
BGP router table entry for 11.0.0.0/24
 Versions:
  Process          bRIB/RIB  SendTblVer
  Speaker                 2           2
 
 Paths: (3 available, best #1)
   Advertised to update-groups (with more than one peer):
     0.1
   Advertised to peers (in unique update groups):
     10.4.101.1
   Received by speaker 0
   Local
     0.0.0.0 from 0.0.0.0 (10.4.0.1)
       Origin IGP, metric 0, localpref 100, weight 32768, valid, local, best 
   Received by speaker 0
   2 3 4
     10.4.101.1 from 10.4.101.1 (10.4.101.1)
       Origin IGP, localpref 100, valid, external
   Received by speaker 0
   Local
     10.4.101.2 from 10.4.101.2 (10.4.101.2)
       Origin IGP, localpref 100, valid, internal 

This table describes the significant fields shown in the display.

Table 4. show bgp prefix length Field Descriptions

Field

Description

BGP router table entry

Network that is being displayed.

Versions

List of the network versions in each BGP process.

Process

Name of the BGP process.

bRIB/RIB

Version of the network for sending to the RIB. You can compare this version with the bRIB/RIB version for the process (at the top of show bgp summary ) to verify whether the network has been sent to the RIB.

SendTblVer

Version of the network for advertising to neighbors. This can be compared with the neighbor version to determine whether the network has been advertised to a particular neighbor.

Paths

List of paths for the network (that is, routes to reach the network). The number of paths and the index of the best path are given.

not advertised to any peer

Best path was received with a NO_ADVERTISE community and is not advertised to any neighbor.

not advertised to EBGP peer

Best path was received with a NO_EXPORT community and is not advertised to any eBGP neighbor.

not advertised outside local AS

Best path was received with a LOCAL_AS community and is not advertised to peers outside the local AS.

Advertisements of this net are suppressed by an aggregate

Network is a more-specific prefix of a configured aggregate and has been suppressed. It is not advertised to any neighbors unless they have an unsuppress-map configured.

Advertised to update-groups

List of update-groups to which the net has been advertised. Update-groups that have only one peer are not listed here.

Advertised to peers

List of neighbors to which the net has been advertised to. Neighbors that are in one of the update-groups listed above are not listed separately. Only neighbors that are in unique update-groups are listed.

Received by speaker 0

BGP process where the path originated. This is always “speaker 0” for standalone mode. It will be the speaker-id when BGP is in distributed mode.

AS Path

Autonomous system (AS) path that was received for the path. If the AS path is empty, then “Local” is displayed. This is the case for paths that are locally generated on this router or on a neighboring router within the same AS.

aggregated by

If the path is an aggregate, the router-id of the router that performed the aggregation.

suppressed due to dampening

Path has been suppressed due to the configured path dampening.

history entry

Path is withdrawn, but a copy is kept to store the dampening information.

Received from a RR-client

Path was received from a route reflector client.

received-only

If soft reconfiguration inbound is configured, the path was received but dropped by inbound policy, or was accepted and modified. In either event, the received-only value is a copy of the original, unmodified path.

received & used

If soft reconfiguration inbound is configured, the path was received and accepted by inbound policy, but not modified.

stale

Neighbor from which the path was received is down, and the path is kept and marked as stale to support graceful restart.

<nexthop> from <neighbor> (<router-id>)

Next hop for the path. If the next hop is known by a mechanism outside BGP (for example, for redistributed paths), then 0.0.0.0 is displayed. After the next hop, the neighbor from whom the path was received is displayed, along with the neighbor’s router-id. If the path was locally generated (for example, an aggregate or redistributed path), then 0.0.0.0 is displayed for the neighbor address.

Origin

IGP: the path originated from an IGP.

EGP: the path originated from an EGP.

incomplete: the origin of the path is unknown.

metric

MED value of the path.

localpref

Local preference value. This is used to determine the preferred exit point from the local autonomous system. It is propagated throughout the local autonomous system.

weight

Locally assigned weight (if not 0) of the path. Weight is used in choosing the preferred path to a route. It is not advertised to any neighbor.

valid

Path is valid and can be considered in the best-path calculation.

redistributed

Path is redistributed through a redistribute command.

aggregated

Path is a locally generated aggregate created due to an aggregate-address command.

local

Path is a local network source due to a network command.

internal

Path was received from an iBGP neighbor.

external

Path was received from an eBGP neighbor.

atomic-aggregate

Path was received with the atomic-aggregate flag set. Some path information has been removed through aggregation.

best

Path is the best path for the network and is used for routing and advertised to peers.

multipath

Path is a multipath and is installed into the RIB along with the best path.

Community

List of communities attached to the path.

Extended community

List of extended communities attached to the path.

Originator

Originator of the path within the AS Cluster list if the path is reflected.

AS Cluster list

List of RR clusters the path has passed through if the path is reflected

Dampinfo

Penalty and reuse information if the path is dampened.

penalty

Current penalty for the path.

flapped

Number of times the path has flapped and the time since the first flap.

reuse in

Time until the path is re-used (undampened).

half life

Configured half-life for the path.

suppress value

Penalty at which the path is suppressed.

reuse value

Penalty at which the path is re-used.

Maximum suppress time

Maximum length of time for which the path can be suppressed.

The following is sample output from the show bgp command with the ip-address/prefix-length detail options:


RP/0/RSP0/CPU0:router# show bgp 51.0.0.0/24 detail 

Sat Mar 14 00:37:14.109 PST PDT

BGP routing table entry for 51.0.0.0/24

Versions:

  Process           bRIB/RIB  SendTblVer

  Speaker                  3           3

    Flags: 0x3e1000, label_retention: not enabled

Last Modified: Mar 13 19:32:17.976 for 05:04:56

Paths: (1 available, best #1)

  Advertised to update-groups (with more than one peer):

    0.3 0.4 0.7 0.8 

  Advertised to peers (in unique update groups):

    201.48.20.1 

  Path #1: Received by speaker 0

  Flags: 0x1000003

  200 201

    213.0.0.6 from 213.0.0.6 (200.200.3.1)

      Origin IGP, localpref 100, valid, external, best

This is sample output to explain 'import suspect' state and 'import-suspect' field in show bgp command output:

RP/0/RSP0/CPU0:router#show bgp vpnv4 unicast rd 11:111 100.16.11.0/24
BGP routing table entry for 100.16.11.0/24, Route Distinguisher: 11:111
Versions:
  Process           bRIB/RIB  SendTblVer
  Speaker            1834195     1834195
Paths: (2 available, best #1)
  Advertised to update-groups (with more than one peer):
    0.1 
  Path #1: Received by speaker 0
  11
    1:16.16.16.16 (metric 30) from 55.55.55.55 (16.16.16.16)
      Received Label 19602
      Origin incomplete, localpref 100, valid, internal, best, import-candidate, not-in-vrf, import suspect
      Extended community: RT:11:11 
      Originator: 16.16.16.16, Cluster list: 55.55.55.55
  Path #2: Received by speaker 0
  11
    1:16.16.16.16 (metric 30) from 88.88.88.88 (16.16.16.16)
      Received Label 19602
      Origin incomplete, localpref 100, valid, internal, not-in-vrf, import suspect
      Extended community: RT:11:11 
      Originator: 16.16.16.16, Cluster list: 88.88.88.88

The show bgp command output displays 'import suspect' when potential import oscillation has been detected for the prefix. Import of such a prefix is not affected. However, import of the prefix can be dampened in future if the oscillation continues. If the oscillation stops during the next import run, the prefix will no longer be marked 'import supect'.

The output of show bgp {vpnv4 | vpnv6} unicast rd command may display the optional BGP attribute not-in-vrf. If a path in a VPNvX net is marked as not-in-vrf, it may be due to any of the following conditions:
  • The RD of the VPNvX net is not the same as any of the RDs configured for VRFs on the router.

  • The RD of the VPNvX net is the same as the RD configured for a specific VRF on the router, but the path is not imported to the specified VRF. For example, the route-targets attached to the path do not match any of the import route-target [as-number:nn | ip-address:nn] configured for VRF, vrf_1 .

If the not-in-vrf net is set, it indicates that the path does not belong to the VRF.

Examples

This is a sample output of show bgp command that displays large-communities in the list of attributes:
RP/0/RSP0/CPU0:router#show bgp 3.3.3.3/32
Thu Mar 23 14:36:15.301 PDT
BGP routing table entry for 3.3.3.3/32
Versions:
  Process           bRIB/RIB  SendTblVer
  Speaker                 42          42
Last Modified: Mar 22 20:04:46.000 for 18:31:30
Paths: (1 available, best #1)
  Advertised to peers (in unique update groups):
    11.11.11.5
  Path #1: Received by speaker 0
  Advertised to peers (in unique update groups):
    11.11.11.5
  Local
    10.10.10.3 from 10.10.10.3 (3.3.3.3)
      Origin incomplete, metric 0, localpref 94, valid, internal, best, group-best
      Received Path ID 0, Local Path ID 0, version 42
      Community: 258:259 260:261 262:263 264:265
     Large Community: 1:2:3 5:6:7 4123456789:4123456780:4123456788

Examples

This is a sample output of show bgp command that displays the net as an aggregate-contributor after applying the aggregate routing policy:

Router#show bgp 250.2.2.1/32
Thu Dec  1 10:12:15.374 EST
BGP routing table entry for 250.2.2.1/32
Versions:
  Process           bRIB/RIB  SendTblVer
  Speaker                 247          247
Last Modified: Dec  1 09:00:20.000 for 01:11:55
Paths: (1 available, best #1)
Net is an aggregate-contributor
  Advertised IPv4 Unicast paths to update-groups (with more than one peer):
    0.2 
  Advertised IPv4 Unicast paths to peers (in unique update groups):
    192.168.0.5     10.10.10.1      
  Path #1: Received by speaker 0
  Advertised IPv4 Unicast paths to update-groups (with more than one peer):
    0.2 
  Advertised IPv4 Unicast paths to peers (in unique update groups):
    192.168.0.5     10.10.10.1      
  105
    13.0.1.1 from 13.0.1.1 (13.0.1.1)
      Origin IGP, localpref 100, valid, external, best, group-best
      Received Path ID 0, Local Path ID 1, version 247
      Community: 20:20
      Origin-AS validity: (disabled)

This is a sample output of show bgp command that displays all aggregate-contributors of the BGP address 250.2.2.0/24:


Router1#show bgp 250.2.2.0/24 aggregate-contributors 
Thu Dec  1 10:15:20.393 EST
BGP router identifier 192.168.0.2, local AS number 100
BGP generic scan interval 60 secs
Non-stop routing is enabled
BGP table state: Active
Table ID: 0xe0000000   RD version: 247
BGP main routing table version 247
BGP NSR Initial initsync version 22 (Reached)
BGP NSR/ISSU Sync-Group versions 247/0
BGP scan interval 60 secs

Status codes: s suppressed, d damped, h history, * valid, > best
              i - internal, r RIB-failure, S stale, N Nexthop-discard
Origin codes: i - IGP, e - EGP, ? - incomplete
   Network            Next Hop            Metric LocPrf Weight Path
*> 250.2.2.1/32       13.0.1.1                               0 105 i
*> 250.2.2.2/32       13.0.2.1                               0 105 i
*> 250.2.2.3/32       13.0.3.1                               0 101 i
*> 250.2.2.4/32       13.0.4.1                               0 101 i
*> 250.2.2.5/32       13.0.5.1                               0 102 i
*> 250.2.2.6/32       13.0.6.1                               0 102 i
*> 250.2.2.7/32       13.0.7.1                               0 103 i
*> 250.2.2.8/32       13.0.8.1                               0 103 i
*> 250.2.2.9/32       13.0.9.1                               0 104 i
*> 250.2.2.10/32      13.0.10.1                              0 104 i
*> 250.2.2.11/32      0.0.0.0                  0         32768 ?
*> 250.2.2.12/32      0.0.0.0                  0         32768 ?
*> 250.2.2.13/32      0.0.0.0                  0         32768 ?

Processed 13 prefixes, 13 paths

After setting an aggregate-contributor to a specific aggregate-address, you can then set the BGP attributes (for example, cost community, next-hop, BGP multiple exit discriminator) to the aggregate-contributor, and then apply the inbound policy and the outbound policy to the neighbors. This is a sample output of show bgp command that displays these details:

For inbound policy:


Router1#show bgp 250.2.2.1/32
Thu Dec  1 10:12:15.374 EST
BGP routing table entry for 250.2.2.1/32
Versions:
  Process           bRIB/RIB  SendTblVer
  Speaker                 247          247
Last Modified: Dec  1 09:00:20.000 for 01:11:55
Paths: (1 available, best #1)
Net is an aggregate-contributor
  Advertised IPv4 Unicast paths to update-groups (with more than one peer):
    0.2 
  Advertised IPv4 Unicast paths to peers (in unique update groups):
    192.168.0.5     10.10.10.1      
  Path #1: Received by speaker 0
  Advertised IPv4 Unicast paths to update-groups (with more than one peer):
    0.2 
  Advertised IPv4 Unicast paths to peers (in unique update groups):
    192.168.0.5     10.10.10.1      
  105
    13.0.1.1 from 13.0.1.1 (13.0.1.1)
      Origin IGP, localpref 100, valid, external, best, group-best
      Received Path ID 0, Local Path ID 1, version 247
      Community: 20:20
      Origin-AS validity: (disabled)
For outbound policy:

Router2# show bgp ipv4 u 250.2.2.4/32
Thu Dec  8 15:08:40.672 EST
BGP routing table entry for 250.2.2.4/32
Versions:
  Process           bRIB/RIB  SendTblVer
  Speaker                 174          174
Last Modified: Dec  8 15:06:53.000 for 00:01:47
Paths: (1 available, best #1)
  Advertised IPv4 Unicast paths to update-groups (with more than one peer):
    0.1 
  Advertised IPv4 Unicast paths to peers (in unique update groups):
    12.12.12.4      
  Path #1: Received by speaker 0
  Advertised IPv4 Unicast paths to update-groups (with more than one peer):
    0.1 
  Advertised IPv4 Unicast paths to peers (in unique update groups):
    12.12.12.4      
  101, (Received from a RR-client)
    13.0.4.1 (metric 2) from 192.168.0.2 (192.168.0.2)
      OC-RIB Attribute-Index 0
      Origin IGP, localpref 100, valid, internal, best, group-best
      Received Path ID 1, Local Path ID 1, version 174
      Extended community: RT:200:200       

show bgp large-community

To display routes that have the specified Border Gateway Protocol (BGP) large-communities, use the show bgp large-community command in EXEC mode.

show bgp [ ipv4 { unicast | multicast | labeled-unicast | all | tunnel } ] large-community large-community-list [ exact-match ]

show bgp [ ipv6 { unicast } ] large-community large-community-list [ exact-match ]

show bgp [ all { unicast | multicast | labeled-unicast | all | tunnel } ] large-community large-community-list [ exact-match ]

show bgp [ vpnv4 unicast [ rd rd-address ] ] large-community large-community-list [ exact-match ]

show bgp [ vrf { vrf-name | all } [ ipv4 | { unicast | labeled-unicast } | ipv6 unicast ] ] large-community large-community-list [ exact-match ]

Syntax Description

ipv4

(Optional) Specifies IP Version 4 address prefixes.

unicast

(Optional) Specifies unicast address prefixes.

multicast

(Optional) Specifies multicast address prefixes.

labeled-unicast

(Optional) Specifies labeled unicast address prefixes.

all

(Optional) For subaddress families, specifies prefixes for all subaddress families.

tunnel

(Optional) Specifies tunnel address prefixes.

ipv6

(Optional) Specifies IP Version 6 address prefixes.

all

(Optional) For address family, specifies prefixes for all address families.

vpnv4 unicast

(Optional) Specifies VPNv4 unicast address families.

rd rd-address

(Optional) Displays routes with a specific route distinguisher.

vrf

(Optional) Specifies VPN routing and forwarding (VRF) instance.

vrf-name

(Optional) Name of a VRF.

all

(Optional) For VRF, specifies all VRFs.

ipv4 { unicast | labeled-unicast }

(Optional) For VRF, specifies IPv4 unicast or labeled-unicast address families.

ipv6 unicast

(Optional) For VRF, specifies IPv6 unicast address families.

large-community

Specifies that only routes with large-communities specified by large-community-list is displayed.

large-community-list

They are specified as three non negative decimal integers separated by colons. For example, 1:2:3. Each integer is stored in 32 bits. The possible range for each integer is 0 to 4294967295.

exact-match

(Optional) Displays those routes that have large- communities exactly matching the specified large-communities.

Command Default

None.

Command Modes

EXEC

Command History

Release

Modification

Release 6.3.1

This command was introduced.

Usage Guidelines

To use this command, you must be in a user group associated with a task group that includes appropriate task IDs. If the user group assignment is preventing you from using a command, contact your AAA administrator for assistance.

Use the exact-match keyword to display only those routes with a set of large communities exactly matching the list of specified large communities. If you omit the exact-match keyword, those routes containing at least the specified large communities are displayed.

Task ID

Task ID

Operations

bgp

read

Examples

The following is sample output from the show bgp large-community command in EXEC mode:


RP/0/RSP0/CPU0:router#show bgp large-community 1:2:3 5:6:7
Thu Mar 23 14:40:33.597 PDT
BGP router identifier 4.4.4.4, local AS number 3
BGP generic scan interval 60 secs
Non-stop routing is enabled
BGP table state: Active
Table ID: 0xe0000000   RD version: 66
BGP main routing table version 66
BGP NSR Initial initsync version 3 (Reached)
BGP NSR/ISSU Sync-Group versions 66/0
BGP scan interval 60 secs

Status codes: s suppressed, d damped, h history, * valid, > best
              i - internal, r RIB-failure, S stale, N Nexthop-discard
Origin codes: i - IGP, e - EGP, ? - incomplete
   Network            Next Hop            Metric LocPrf Weight Path
*>i3.3.3.3/32         10.10.10.3               0     94      0 ?
*> 5.5.5.5/32         11.11.11.5               0             0 5 ?

In the above example, the routes 3.3.3.3/32 and 5.5.5.5/32 contain both the large communities 1:2:3 and 5:6:7 and some additional large communities.

RP/0/RSP0/CPU0:router#show bgp large-community 1:2:3 5:6:7 exact-match
Thu Mar 23 14:50:28.155 PDT

In the above example, no routes are displayed because the available routes have some additional large-communities, so the match is not exact.

show bgp bmp

To display Border Gateway Protocol (BGP) Monitoring Protocol (BMP) information, use the show bgp bmp command in EXEC mode.

show bgp bmp { server server-id [detail] | summary }

show bgp bmp server <server-id> [[vrf <vrf> ] neighbor <neighbor-ip> ] [detail | wide]

show bgp bmp [mode {inbound post-policy | local-rib | all}] [<afi> <safi>] table [rd <rd> | vrf <vrf-name> ] [neighbor <address> | version begin <begin-version> end <end-version> | prefix <prefix-length> | summary]

show bgp bmp [mode {inbound post-policy | local-rib | all}] vrf <vrf-name> [<afi> <safi>] table [neighbor <address> | version begin <begin-version> end <end-version> | prefix] <prefix-length>

show bgp bmp mode local-rib [vrf <vrf-name> ] emulated-peers <emulated-peer-address> [detail | performance-statistics | brief]

show bgp bmp mode local-rib [vrf <vrf-name> ] internal update-group <update-group-index> [summary] | emulated-peer <emulated-peer-address> ] [performance-statistics]

show bgp bmp mode local-rib internal update out [emulated-peer emulated-peer-address ]

show bgp bmp [mode {inbound post-policy | local-rib | all}] [<afi> <safi>] convergence

Syntax Description

server server-id

Displays information about BMP server as specified by the server-id variable.

detail

(Optional) Displays detailed BMP server information.

summary

Displays summary information about all the configured BMP servers.

wide

Wide tabular formatting of the BMP neighbor details.

table

Show prefixes and paths from the Inbound Post-Policy and Local RIB tables.

ipv4

IP Version 4 address prefixes

ipv6

IP Version 6 address prefixes

vpnv4

VPN version 4 address family

vpnv6

VPN version 6 address family

unicast

Unicast sub-address family

labeled-unicast

Labeled Unicast sub-address family

all

  • When used as address family, the command is executed for all address families.

  • When used as sub-address family, the command is executed for all sub-address families.

neighbor

Output of the command is filtered for a specific neighbor.

mode

Execute the command for the specified BMP monitoring mode.

rd

Output of the command is filtered for a specific route distinguisher.

vrf

Output of the command is filtered for a specific VRF.

summary

Show only the summary information

begin

Begin version of a version range that is used for filtering the prefixes within that range.

end

End version of a version range that is used for filtering the prefixes within that range.

emulated-peers

BMP Local-RIB emulated peers

brief

Display the output in brief tabular format.

detail

Display detailed information

performance-statistics

Display performance statistics

Command Default

No default behavior or values

Command Modes

EXEC

Command History

Release

Modification

Release 24.1.1

The option to specify AFIs and SAFIs after the table keyword is completely removed from Cisco IOS XR Release 24.1.1.

Release 7.5.4

The syntax of the show bgp bmp table vpnv4 unicast command was modified. Specifying AFI and SAFI after the table keyword option is not supported from Release 7.5.4. So, it is recommended to specify AFIs and SAFIs before the table keyword.

Release 7.5.4

This command was modified with Inbound Post-Policy and Local RIB view.

Release 5.2.2

This command was introduced.

Usage Guidelines

To use this command, you must be in a user group associated with a task group that includes appropriate task IDs. If the user group assignment is preventing you from using a command, contact your AAA administrator for assistance.

Inbound post-policy and Local-RIB tables
  • If AFI is not specified while running the commands, the default IPv4 Unicast AFI is considered.

  • If mode is not specified, the default mode inbound post-policy is considered.

  • Specifying AFIs and SAFIs after the table keyword option is not supported from Release 7.5.4. So, you can specify the same before the table keyword as follows:

    show bgp bmp [mode <bmp-mode>] [<afi> <safi>] table

  • Starting from Release 7.5.4, if AFIs or SAFIs are specified before and after the table keyword, the AFIs or SAFIs after the table keyword are considered.

  • The following filters are only available for vpnv4 unicast and vpnv6 unicast AFIs:

    • rd

    • vrf

  • The begin-version of the bmp table version can be smaller than or greater than or equal to end-version.

Emulated-Peers
  • The keyword brief can be used only when vrf is not specified.

  • The keyword emulated-peers is only applicable for Local-RIB.

BGP BMP Convergence

Use the show bgp bmp convergence  command to see if there is any pending work for BGP to perform. The software checks the following conditions to determine whether the specified address family has converged:

  • Table copy is not pending (applicable only in case of Inbound Post-Policy)

  • Update generation is not pending

  • All BMP servers that are not administratively shutdown are Up state

  • All BMP servers have caught up to the table version

  • Write queues of all BMP servers are empty

If all the above conditions are true, the address family is considered converged.

Task ID

Task ID

Operations

bgp

read

Examples

The following example shows sample output from the show bgp bmp command when the summary keyword is used:


RP/0/RSP0/CPU0:router# show bgp bmp summary
ID   Host                Port     State   Time        NBRs
 1   10.0.101.1          16666    ESTAB   00:29:52    1   
 2   10.0.101.2          16667    ESTAB   00:29:52    0   
 3   fed0::1001          26666    ESTAB   00:29:52    0   
 4   fed0::1002          26667    ESTAB   00:29:52    0   
 5   10.0.101.1          16666    ESTAB   00:21:49    0   
 6   10.0.101.1          16666    ESTAB   00:29:52    0   
 7   fed0::1001          26666    ESTAB   00:29:52    0   
 8   fed0::1001          26666    ESTAB   00:29:52    0 

The following example shows sample output from the show bgp bmp command when the server keyword, with server ID as 4, is used:


RP/0/RSP0/CPU0:router# show bgp bmp server 4
BMP server 4
Host 10.0.101.1 Port 16666
Connected for 00:25:07
Precedence:  internet
BGP neighbors: 1
VRF: - (0x60000000)
Update Source: 9.9.9.9 (Lo9)
Update Source Vrf ID: 0x60000000

Message Stats:
Total messages sent: 60
         INITIATION: 1
        TERMINATION: 0
       STATS-REPORT: 0
  PER-PEER messages: 59

  Neighbor 20.0.101.11
Messages pending: 0
Messages sent: 59
      PEER-UP: 1
    PEER-DOWN: 0
    ROUTE-MON: 58

The show bgp bmp summary command displays BMP server summary information such as server address and port number, server state, time for which server is up and the number of neighbors the server is monitoring.


Router#show bgp bmp summary 

ID   Host                 Port     State   Time        NBRs 

 1   12.1.2.1             16001    ESTAB   01:35:05    14   

 2   12.2.2.1             16002    ESTAB   01:30:47    11 
The show bgp bmp server <ID> command displays BMP server details, BMP update modes exported to the BMP server, configured (or default) values specific to the update modes, update generation details, TCP connection status, BMP server specific message statistics, and statistics of the neighbors monitored by the BMP server.

Router#show bgp bmp server 1      
BMP server 1 
Host 12.1.2.1 Port 16001 
Connected for 01:19:20 
Last Disconnect event received : 00:00:00 
Precedence:  internet 
BGP neighbors: 14 
VRF: - (0x60000000) 
Update Source: - (-) 
Update Source Vrf ID: 0x0 
Update Mode                       : In-Post-Policy, Local-RIB 
  In-Post-Policy 
   Advertisement interval         : 15 secs 
   Scanner interval               : 60 secs 
  Local-RIB 
   Advertisement interval         : 15 secs 
   Scanner interval 
     Global                       : 60 secs 
     IPv4 Unicast                 : 60 secs 
     VPNv4 Unicast                : 60 secs 
     IPv6 Unicast                 : 60 secs 
     VPNv6 Unicast                : 60 secs 
Flapping Delay                    : 300 secs 
Initial Delay                     : 0 secs 
Initial Refresh Delay             : 1 secs 
Initial Refresh Spread            : 1 secs 
Stats Reporting Period            : 0 secs 
Queue Route Mon Msg buffer limit  : 133693 KB (Current Server Up Count: 2) 
Queue Route Mon Msg buffer usage  : 0 B 
Queue write pulse sent            : Feb  6 12:32:13.676, not set (all) 
Queue write pulse received        : Feb  6 12:32:13.676 
Update Generation in Progress     : No 
Reset Walk in Progress            : No 
IPv4 Unicast         
  Version                     : 11 
  Init EOR Version            : 11 
  Init EOR Pending count      : 0 
  Update Generation  
    Last Run                   : Feb  6 12:32:04.615, Count 1 
    Walk Currently Stalled     : No, Last Stalled : not set, Count 0 
VPNv4 Unicast        
  Version                     : 14 
  Init EOR Version            : 14 
  Init EOR Pending count      : 0 
  Update Generation  
    Last Run                   : Feb  6 12:32:04.615, Count 1 
    Walk Currently Stalled     : No, Last Stalled : not set, Count 0 
IPv6 Unicast         
  Version                     : 12 
  Init EOR Version            : 12 
  Init EOR Pending count      : 0 
  Update Generation  
    Last Run                   : Feb  6 12:32:04.615, Count 1 
    Walk Currently Stalled     : No, Last Stalled : not set, Count 0 
VPNv6 Unicast        
  Version                     : 16 
  Init EOR Version            : 16 
  Init EOR Pending count      : 0 
  Update Generation  
    Last Run                   : Feb  6 12:32:04.615, Count 1 
    Walk Currently Stalled     : No, Last Stalled : not set, Count 0 
  
TCP:  
  Last message sent: Feb  6 12:32:04.615, Status: No Pending Data 
  Last write pulse received: Feb  6 12:32:04.863, Waiting: FALSE 

Message Stats: 
Total msgs dropped   : 0 
Total msgs pending   : 0, Max: 19 at Feb  6 12:32:04.615 
Total messages sent  : 100 
Total bytes sent     : 13148, Time spent: 0.000 secs 
  INITIATION         : 1 
  TERMINATION        : 0 
  STATS-REPORT       : 0 
  PER-PEER messages  : 99  
  ROUTE-MON messages : 88 
   EOR messages      : 30 
   Update messages   : 58 (Prefixes: 41, Err: 0)  
Update gen time spent: 0.000 secs 

Neighbor 12.100.0.1 (vrf foo) 
  Messages pending : 0 
  Messages dropped : 0 
  Messages sent    : 5 
    PEER-UP        : 1 
    PEER-DOWN      : 0 
    ROUTE-MON      : 4 
      EOR          : 2 
      Update       : 2 (Prefixes: 4, Err: 0)  

Neighbor 12:100::1 (vrf foo) 
  Messages pending : 0 
  Messages dropped : 0 
  Messages sent    : 3 
    PEER-UP        : 1 
    PEER-DOWN      : 0 
    ROUTE-MON      : 2 
      EOR          : 1 
      Update       : 1 (Prefixes: 2, Err: 0)  

Neighbor 12.101.0.1 (vrf bar) 
  Messages pending : 0 
  Messages dropped : 0 
  Messages sent    : 5 
    PEER-UP        : 1 
    PEER-DOWN      : 0 
    ROUTE-MON      : 4 
      EOR          : 2 
      Update       : 2 (Prefixes: 4, Err: 0)  

Neighbor 12.0.0.1 (vrf default) 
  Messages pending : 0 
  Messages dropped : 0 
  Messages sent    : 7 
    PEER-UP        : 1 
    PEER-DOWN      : 0 
    ROUTE-MON      : 6 
      EOR          : 4 
      Update       : 2 (Prefixes: 4, Err: 0) 
The show bgp bmp server <ID> command with the wide command displays the same information as show bgp bmp server <ID> except that it displays the neighbor information in a wide tabular format.

Router#show bgp bmp server 1 wide 
BMP server 1 
Host 12.1.2.1 Port 16001 
Connected for 01:00:49 
Last Disconnect event received : 00:00:00 
Precedence:  internet 
BGP neighbors: 14 
VRF: - (0x60000000) 
Update Source: - (-) 
Update Source Vrf ID: 0x0 
.. 
<truncated> .. 

Message Stats: 
Total msgs dropped   : 0 
Total msgs pending   : 0, Max: 19 at Feb  6 12:32:04.615 
Total messages sent  : 100 
Total bytes sent     : 13148, Time spent: 0.000 secs 
  INITIATION         : 1 
  TERMINATION        : 0 
  STATS-REPORT       : 0 
  PER-PEER messages  : 99 
  ROUTE-MON messages : 88 
   EOR messages      : 30 
   Update messages   : 58 (Prefixes: 41, Err: 0)  

Update gen time spent: 0.000 secs 
 Legend: 
  MsgPend   : Messages yet to be sent 
  MsgDrop   : Messages dropped 
  MsgSent   : Messages sent 
  PeerUp    : Peer-Up messages sent 
  PeerDown  : Peer-Down messages sent 
  RouteMon  : Router-Monitoring messages sent 
  EOR       : EOR messages sent 
  Update    : Update messages sent 
  UpdPaths  : Paths advertised in Update messages 
  UpdErrs   : Paths not advertised in Update messages 
  Withdraw  : Withdraw messages sent 
  WdrwPaths : Paths advertised in Withdraw messages 
  WdrwErrs  : Paths not advertised in Withdraw messages 

Neighbor                 VRF                     MsgPend    MsgDrop    MsgSent     PeerUp   PeerDown   RouteMon        EOR     Update   UpdPaths    UpdErrs   Withdraw  WdrwPaths   WdrwErrs 

12.100.0.1               foo                           0          0          5          1          0          4          2          2          4          0          0          0          0 
12:100::1                foo                           0          0          3          1          0          2          1          1          2          0          0          0          0 
12.101.0.1               bar                           0          0          5          1          0          4          2          2          4          0          0          0          0 
12.0.0.1                 default                       0          0          7          1          0          6          4          2          4          0          0          0          0 
The show bgp bmp server <ID> command displays the advertisement-interval time that you configured.

Router#show bgp bmp server 1 
BMP server 1 
Host 12.1.2.1 Port 16001 
Connected for 04:22:39 
Last Disconnect event received : 00:00:00 
Precedence:  internet 
BGP neighbors: 10 
VRF: - (0x60000000) 
Update Source: - (-) 
Update Source Vrf ID: 0x0 
Update Mode                       : In-Post-Policy, Local-RIB 
  In-Post-Policy 
   Advertisement interval         : 15 secs 
   Scanner interval               : 60 secs 
  Local-RIB 
   Advertisement interval         : 15 secs 
   Scanner interval 
     Global                       : 60 secs 
     IPv4 Unicast                 : 60 secs 
     VPNv4 Unicast                : 60 secs 
     IPv6 Unicast                 : 60 secs 
     VPNv6 Unicast                : 60 secs 
Flapping Delay                    : 300 secs 
Initial Delay                     : 0 secs 
Initial Refresh Delay             : 1 secs 
Initial Refresh Spread            : 1 secs 
Stats Reporting Period            : 0 secs 
Queue Route Mon Msg buffer limit  : 133693 KB (Current Server Up Count: 2) 
Queue Route Mon Msg buffer usage  : 0 B 
Queue write pulse sent            : Dec 16 15:19:36.755, not set (all) 
-----More------ 
The show bgp bmp server <ID> command displays the BMP scan time that you configured.

Router#show bgp bmp server 1  
BMP server 1 
Host 12.1.2.1 Port 16001 
Connected for 04:22:39 
Last Disconnect event received : 00:00:00 
Precedence:  internet 
BGP neighbors: 10 
VRF: - (0x60000000) 
Update Source: - (-) 
Update Source Vrf ID: 0x0 
Update Mode                       : In-Post-Policy, Local-RIB 
  In-Post-Policy 
   Advertisement interval         : 15 secs 
   Scanner interval               : 60 secs 
  Local-RIB 
   Advertisement interval         : 15 secs 
   Scanner interval 
     Global                       : 60 secs 
     IPv4 Unicast                 : 60 secs 
     VPNv4 Unicast                : 60 secs 
     IPv6 Unicast                 : 60 secs 
     VPNv6 Unicast                : 60 secs 
Flapping Delay                    : 300 secs 
Initial Delay                     : 0 secs 
Initial Refresh Delay             : 1 secs 
Initial Refresh Spread            : 1 secs 
Stats Reporting Period            : 0 secs 
Queue Route Mon Msg buffer limit  : 133693 KB (Current Server Up Count: 2) 
Queue Route Mon Msg buffer usage  : 0 B 
Queue write pulse sent            : Dec 16 15:19:36.755, not set (all) 
-----More------ 

 
The show bgp bmp table command displays summary information of prefixes and paths learnt from the monitored BGP neighbors.

Router#show bgp bmp table 
---------------------------------------------- 
Route monitoring mode : Inbound Post-Policy 
---------------------------------------------- 
Address-Family : IPv4 Unicast 
------------------------------------ 
BMP prefix table version : 11 
Number of prefixes: 5, paths: 9 

   Network            Version    Neighbor         Bit Map    Flags  Path ID 
   100.1.1.3/32       7          23.0.1.3         0x03/0xfc  0x02   1 
   100.1.2.3/32       11         23:0:2::3        0x03/0xfc  0x02   1 
    100.1.2.3/32      11         23.0.2.3         0x03/0xfc  0x02   1 
   100.1.3.3/32       10         23.0.2.3         0x03/0xfc  0x03   1 
    100.1.3.3/32      10         23.0.1.3         0x03/0xfc  0x03   1 
   200.1.1.1/32       5          12::1            0x03/0xfc  0x00    
    200.1.1.1/32      5          12.0.0.1         0x03/0xfc  0x00    
   200.2.1.1/32       6          12::1            0x03/0xfc  0x00    
    200.2.1.1/32      6          12.0.0.1         0x03/0xfc  0x00 

If a prefix has more than one path, the prefix is printed multiple times, once for each path, except that for the second paths onwards, the prefix is indented by one space as visible in case of the prefix 100.1.2.3 in the above table. This prefix has a path via neighbor 23:0:2::3 and 23.0.2.3.

The Bit Map displayed above has 2 bitmap values. The first value is the bitmap of BMP servers to which the prefix is advertised and the second value is the bitmap of BMP servers to which the prefix is yet to be advertized. Bit n in both the bitmaps (where n is 0 to 7) represent Server m where m = n + 1. For example, bit 3 represents Server 4.

If the advertized bitmap is 0x03 (00000011b) it means that the prefix is advertized to servers 1 and 2. Simularly, if the yet to be advertized bitmap is 0xfc, then it means that the prefix is yet to be advertized to servers 3, 4, 5, 6, 7 and 8. However if any of these servers is not configured, then the corresponding bits can be ignored.

You can run the show bgp bmp table command as follows:

  • For a specific address family identifiers (AFI) or sub-address family identifiers (SAFI). For example, show bgp bmp vpnv4 unicast table, or a specific BMP monitoring mode, for example show bgp bmp mode inbound post-policy table.

  • If you do not specify the AFI and SAFI, the command displays information for the default AFI IPv4 Unicast.

  • If you do not specify the BMP monitoring mode, the command displays information for the default mode Inbound Post-Policy.

    For example, if you run the command show bgp bmp table where the BMP monitoring mode or the AFI are not specified, the command displays information for Inbound Post-Policy BMP mode IPv4 Unicast AFI.

  • For all BMP monitoring modes (mode all) and/or all AFIs (all all).

The following example shows a brief summary of the prefixes for BMP mode Local-RIB and IPv4 Unicast AFI:

Router#show bgp bmp mode local-rib table 
---------------------------------------------- 
Route monitoring mode : Local-RIB 
---------------------------------------------- 
Address-Family : IPv4 Unicast 
------------------------------------ 
BGP main routing table version : 9 
Number of prefixes: 4, paths: 4 

   Network            Next Hop            Metric LocPrf Path 
   100.1.1.3/32       23.0.1.3                 0    100 ? 
   100.1.2.3/32       23.0.2.3                 0    100 ? 
   200.1.1.1/32       12.0.0.1                 0        1 ? 
   200.2.1.1/32       12.0.0.1                 0        1 ? 
Processed 4 prefixes, 4 paths 

 
The following example shows a brief summary of the prefixes for BMP mode Inbound Post-Policy and VPNv4 Unicast AFI:

Router#show bgp bmp vpnv4 unicast table 

---------------------------------------------- 
Route monitoring mode : Inbound Post-Policy 
---------------------------------------------- 
Address-Family : VPNv4 Unicast 
------------------------------------ 
BMP prefix table version : 14 
Number of RDs: 3, prefixes: 6, paths: 10 
   Network            Version    Neighbor         Bit Map    Flags  Path ID 
Route Distinguisher: 2:100 (default for vrf foo), Version 6 
   200.1.1.1/32       5          12.100.0.1       0x03/0xfc  0x00    
   200.2.1.1/32       6          12.100.0.1       0x03/0xfc  0x00    
Route Distinguisher: 2:101 (default for vrf bar), Version 8 
   200.1.1.1/32       7          12.101.0.1       0x03/0xfc  0x00    
   200.2.1.1/32       8          12.101.0.1       0x03/0xfc  0x00    
Route Distinguisher: 2:200, Version 14 
   200.1.1.4/32       14         23.0.1.3         0x03/0xfc  0x02    
    200.1.1.4/32      14         23.0.2.3         0x03/0xfc  0x02    
    200.1.1.4/32      14         23:0:2::3        0x03/0xfc  0x02    
   200.2.1.4/32       13         23.0.1.3         0x03/0xfc  0x02    
    200.2.1.4/32      13         23.0.2.3         0x03/0xfc  0x02    
    200.2.1.4/32      13         23:0:2::3        0x03/0xfc  0x02    

Processed 6 prefixes, 10 paths 
The following example shows a brief summary of the prefixes for all BMP modes and all AFI:

Router#show bgp bmp mode all all all table 

---------------------------------------------- 
Route monitoring mode : Inbound Post-Policy 
---------------------------------------------- 
Address-Family : VPNv4 Unicast 
------------------------------------ 
BMP prefix table version : 14 
Number of RDs: 3, prefixes: 6, paths: 10 

   Network            Version    Neighbor         Bit Map    Flags  Path ID 
Route Distinguisher: 2:100 (default for vrf foo), Version 6 
   200.1.1.1/32       5          12.100.0.1       0x03/0xfc  0x00    
   200.2.1.1/32       6          12.100.0.1       0x03/0xfc  0x00    
Route Distinguisher: 2:101 (default for vrf bar), Version 8 
   200.1.1.1/32       7          12.101.0.1       0x03/0xfc  0x00    
   200.2.1.1/32       8          12.101.0.1       0x03/0xfc  0x00    
Route Distinguisher: 2:200, Version 14 
   200.1.1.4/32       14         23.0.1.3         0x03/0xfc  0x02    
    200.1.1.4/32      14         23.0.2.3         0x03/0xfc  0x02    
    200.1.1.4/32      14         23:0:2::3        0x03/0xfc  0x02    
   200.2.1.4/32       13         23.0.1.3         0x03/0xfc  0x02    
    200.2.1.4/32      13         23.0.2.3         0x03/0xfc  0x02    
    200.2.1.4/32      13         23:0:2::3        0x03/0xfc  0x02    

Processed 6 prefixes, 10 paths 
Address-Family : VPNv6 Unicast 
------------------------------------ 
BMP prefix table version : 16 
Number of RDs: 3, prefixes: 6, paths: 12 

   Network            Version    Neighbor         Bit Map    Flags  Path ID 
Route Distinguisher: 2:100 (default for vrf foo), Version 8 
   200:1:1::1/128     7          12.100.0.1       0x03/0xfc  0x00    
    200:1:1::1/128    7          12:100::1        0x03/0xfc  0x00    
   200:2:1::1/128     8          12.100.0.1       0x03/0xfc  0x00    
    200:2:1::1/128    8          12:100::1        0x03/0xfc  0x00    

Route Distinguisher: 2:101 (default for vrf bar), Version 10 
   200:1:1::1/128     9          12.101.0.1       0x03/0xfc  0x00    
   200:2:1::1/128     10         12.101.0.1       0x03/0xfc  0x00    

Route Distinguisher: 2:200, Version 16 
   200:1:1::4/128     16         23.0.2.3         0x03/0xfc  0x02    
    200:1:1::4/128    16         23:0:2::3        0x03/0xfc  0x02    
    200:1:1::4/128    16         23.0.1.3         0x03/0xfc  0x02    
   200:2:1::4/128     15         23.0.2.3         0x03/0xfc  0x02    
    200:2:1::4/128    15         23:0:2::3        0x03/0xfc  0x02    
    200:2:1::4/128    15         23.0.1.3         0x03/0xfc  0x02    

Processed 6 prefixes, 12 paths 
Address-Family : IPv4 Labeled-unicast 
------------------------------------ 
BMP prefix table version : 11 
Number of prefixes: 5, paths: 9 
   Network            Version    Neighbor         Bit Map    Flags  Path ID 
   100.1.3.3/32       10         23.0.2.3         0x03/0xfc  0x03   1 
    100.1.3.3/32      10         23.0.1.3         0x03/0xfc  0x03   1 

Processed 1 prefixes, 2 paths 
Address-Family : IPv6 Labeled-unicast 
------------------------------------ 
BMP prefix table version : 12 
Number of prefixes: 5, paths: 10 

   Network            Version    Neighbor         Bit Map    Flags  Path ID 
   100:1:3::3/128     12         23:0:2::3        0x03/0xfc  0x03   1 
    100:1:3::3/128    12         23.0.2.3         0x03/0xfc  0x03   1 
    100:1:3::3/128    12         23.0.1.3         0x03/0xfc  0x03   1 

Processed 1 prefixes, 3 paths 
Address-Family : IPv4 Unicast 
------------------------------------ 
BMP prefix table version : 11 
Number of prefixes: 5, paths: 9  

   Network            Version    Neighbor         Bit Map    Flags  Path ID 
   100.1.1.3/32       7          23.0.1.3         0x03/0xfc  0x02   1 
   100.1.2.3/32       11         23:0:2::3        0x03/0xfc  0x02   1 
    100.1.2.3/32      11         23.0.2.3         0x03/0xfc  0x02   1 
   100.1.3.3/32       10         23.0.2.3         0x03/0xfc  0x03   1 
    100.1.3.3/32      10         23.0.1.3         0x03/0xfc  0x03   1 
   200.1.1.1/32       5          12::1            0x03/0xfc  0x00    
    200.1.1.1/32      5          12.0.0.1         0x03/0xfc  0x00    
   200.2.1.1/32       6          12::1            0x03/0xfc  0x00    
    200.2.1.1/32      6          12.0.0.1         0x03/0xfc  0x00     

Processed 5 prefixes, 9 paths 
Address-Family : IPv6 Unicast 
------------------------------------ 
BMP prefix table version : 12 
Number of prefixes: 5, paths: 10 

   Network            Version    Neighbor         Bit Map    Flags  Path ID 
   100:1:1::3/128     8          23.0.1.3         0x03/0xfc  0x02   1 
   100:1:2::3/128     9          23:0:2::3        0x03/0xfc  0x02   1 
    100:1:2::3/128    9          23.0.2.3         0x03/0xfc  0x02   1 
   100:1:3::3/128     12         23:0:2::3        0x03/0xfc  0x03   1 
    100:1:3::3/128    12         23.0.2.3         0x03/0xfc  0x03   1 
    100:1:3::3/128    12         23.0.1.3         0x03/0xfc  0x03   1 
   200:1:1::1/128     5          12.0.0.1         0x03/0xfc  0x00    
    200:1:1::1/128    5          12::1            0x03/0xfc  0x00    
   200:2:1::1/128     6          12.0.0.1         0x03/0xfc  0x00    
    200:2:1::1/128    6          12::1            0x03/0xfc  0x00    

Processed 5 prefixes, 10 paths 
---------------------------------------------- 
Route monitoring mode : Local-RIB 
----------------------------------------------
Address-Family : VPNv4 Unicast 
------------------------------------ 
BGP main routing table version : 33 
Number of prefixes: 10, paths: 10 

   Network            Next Hop            Metric LocPrf Path 
Route Distinguisher: 2:100 (default for vrf foo), Version 33 
   200.1.1.1/32       12.100.0.1               0        1 ? 
   200.1.1.4/32       23.0.1.3                 0    100 3 ? 
   200.2.1.1/32       12.100.0.1               0        1 ? 
   200.2.1.4/32       23.0.1.3                 0    100 3 ? 

Route Distinguisher: 2:101 (default for vrf bar), Version 31 
   200.1.1.1/32       12.101.0.1               0        1 ? 
   200.1.1.4/32       23.0.1.3                 0    100 3 ? 
   200.2.1.1/32       12.101.0.1               0        1 ? 
   200.2.1.4/32       23.0.1.3                 0    100 3 ? 

Route Distinguisher: 2:200, Version 27 
   200.1.1.4/32       23.0.1.3                 0    100 3 ? 
   200.2.1.4/32       23.0.1.3                 0    100 3 ? 
Processed 10 prefixes, 10 paths 

Address-Family : VPNv6 Unicast 
------------------------------------ 
BGP main routing table version : 31 

Number of prefixes: 10, paths: 10 

   Network            Next Hop            Metric LocPrf Path 
Route Distinguisher: 2:100 (default for vrf foo), Version 29 
   200:1:1::1/128     12:100::1                0        1 ? 
   200:1:1::4/128     23.0.1.3                 0    100 3 ? 
   200:2:1::1/128     12:100::1                0        1 ? 
   200:2:1::4/128     23.0.1.3                 0    100 3 ? 

Route Distinguisher: 2:101 (default for vrf bar), Version 23 
   200:1:1::1/128     12:101::1                0        1 ? 
   200:1:1::4/128     23.0.1.3                 0    100 3 ? 
   200:2:1::1/128     12:101::1                0        1 ? 
   200:2:1::4/128     23.0.1.3                 0    100 3 ? 

Route Distinguisher: 2:200, Version 31 

   200:1:1::4/128     23.0.1.3                 0    100 3 ? 
   200:2:1::4/128     23.0.1.3                 0    100 3 ? 

Processed 10 prefixes, 10 paths 
Address-Family : IPv4 Labeled-unicast 
------------------------------------ 
BGP main routing table version : 9 
Number of prefixes: 1, paths: 1 

   Network            Next Hop            Metric LocPrf Path 

   100.1.3.3/32       23.0.1.3                 0    100 ? 

Processed 1 prefixes, 1 paths 
Address-Family : IPv6 Labeled-unicast 
------------------------------------ 
BGP main routing table version : 9 
Number of prefixes: 1, paths: 1 

   Network            Next Hop            Metric LocPrf Path 

   100:1:3::3/128     23.0.1.3                 0    100 ? 
Processed 1 prefixes, 1 paths 
Address-Family : IPv4 Unicast 
------------------------------------ 
BGP main routing table version : 9 

Number of prefixes: 4, paths: 4 

   Network            Next Hop            Metric LocPrf Path 
   100.1.1.3/32       23.0.1.3                 0    100 ? 
   100.1.2.3/32       23.0.2.3                 0    100 ? 
   200.1.1.1/32       12.0.0.1                 0        1 ? 
   200.2.1.1/32       12.0.0.1                 0        1 ? 

Processed 4 prefixes, 4 paths 
Address-Family : IPv6 Unicast 
------------------------------------ 
BGP main routing table version : 9 
Number of prefixes: 4, paths: 4 

   Network            Next Hop            Metric LocPrf Path 
   100:1:1::3/128     23:0:1::3                0    100 ? 
   100:1:2::3/128     23:0:2::3                0    100 ? 
   200:1:1::1/128     12::1                    0        1 ? 
   200:2:1::1/128     12::1                    0        1 ? 

Processed 4 prefixes, 4 paths 

Note


Specifying AFI and SAFI after the table keyword option is not supported from Release 7.5.4. So, you can specify the same before the table keyword as follows:
show bgp bmp [mode <bmp-mode>] [<afi> <safi>] table 

The option to specify AFIs and SAFIs after the table keyword is completely removed from Cisco IOS XR Release 24.1.1.


The tables for VPN AFI can be filtered on a Route Distinguisher (RD) or a VRF. The following example shows the table for VPNv4 Unicast without RD or VRF filter.

Router#show bgp bmp vpnv4 unicast table  

---------------------------------------------- 
Route monitoring mode : Inbound Post-Policy 
---------------------------------------------- 
Address-Family : VPNv4 Unicast
------------------------------------ 
BMP prefix table version : 18 

Number of RDs: 3, prefixes: 6, paths: 12 

   Network            Version    Neighbor         Bit Map    Flags  Path ID 
Route Distinguisher: 2:100 (default for vrf foo), Version 14 
   200.1.1.1/32       13         12.100.0.1       0x03/0xfc  0x02    
   200.2.1.1/32       14         12.100.0.1       0x03/0xfc  0x02    

Route Distinguisher: 2:101 (default for vrf bar), Version 16 
   200.1.1.1/32       15         12.101.0.1       0x03/0xfc  0x02    
   200.2.1.1/32       16         12.101.0.1       0x03/0xfc  0x02    

Route Distinguisher: 2:200, Version 18 
   200.1.1.4/32       17         23.0.3.3         0x03/0xfc  0x00    
    200.1.1.4/32      17         23.0.2.3         0x03/0xfc  0x00    
    200.1.1.4/32      17         23.0.4.3         0x03/0xfc  0x00    
    200.1.1.4/32      17         23.0.1.3         0x03/0xfc  0x00    
   200.2.1.4/32       18         23.0.3.3         0x03/0xfc  0x00    
    200.2.1.4/32      18         23.0.2.3         0x03/0xfc  0x00    
    200.2.1.4/32      18         23.0.4.3         0x03/0xfc  0x00    
    200.2.1.4/32      18         23.0.1.3         0x03/0xfc  0x00    
The following example shows the show bgp bmp vpnv4 unicast table filtered on route distinguisher 2:100:

Router#show bgp bmp vpnv4 unicast table rd 2:100 

---------------------------------------------- 
Route monitoring mode : Inbound Post-Policy 
---------------------------------------------- 
Address-Family : VPNv4 Unicast 
------------------------------------ 

BMP prefix table version : 18 
Number of RDs: 3, prefixes: 6, paths: 12 

   Network            Version    Neighbor         Bit Map    Flags  Path ID 
Route Distinguisher: 2:100 (default for vrf foo), Version 14 
   200.1.1.1/32       13         12.100.0.1       0x03/0xfc  0x02    
   200.2.1.1/32       14         12.100.0.1       0x03/0xfc  0x02    

Processed 2 prefixes, 2 paths 
The following example shows the show bgp bmp vpnv4 unicast table filtered on VRF foo:

Router#show bgp bmp vpnv4 unicast table vrf foo  
---------------------------------------------- 
Route monitoring mode : Inbound Post-Policy 
---------------------------------------------- 
Address-Family : VPNv4 Unicast 
------------------------------------ 
BMP prefix table version : 18 

Number of RDs: 3, prefixes: 6, paths: 12 

   Network            Version    Neighbor         Bit Map    Flags  Path ID 
Route Distinguisher: 2:100 (default for vrf foo), Version 14 
   200.1.1.1/32       13         12.100.0.1       0x03/0xfc  0x02    
   200.2.1.1/32       14         12.100.0.1       0x03/0xfc  0x02    

Processed 2 prefixes, 2 paths 

 
When the show bgp bmp table command is run with prefix filter, it displays the details of the specified prefix as shown in the following example:

Router#show bgp bmp table 100.1.1.3/32  
---------------------------------------------- 
Route monitoring mode : Inbound Post-Policy 
---------------------------------------------- 
BGP routing table entry for 100.1.1.3/32 

Versions: 
  Process           Net 
  Speaker           7           
Last Modified: Dec 16 15:05:22.000 for 07:05:14 
    Flags: 0x00000000+0x00000000 
    Net Ptr: 0x7f0bbd65ff34 
Paths: (1 available) 
  Path #1: Received by speaker 0 
  Flags: 0x00000000 
  Local 
  from 23.0.1.3 
  Nexthop: 23.0.1.3 
  Received Path ID: 1 
  Received Local Preference: 100 
  Path Advertised to BMP server(s) - (Map: 0x03): Server ID: 1, 2 
  Path pending send to BMP server(s) - (Map: 0xfc): Server ID: None 
  Path Ptr : 0x7f0bb806f0d0, BGP Path Ptr : 0x7f0bbd915f68 

 
The following examples displays the output of the show bgp bmp table command with the summary keyword. It displays the summary of the table of the specified AFI or SAFI, or the default AFI or SAFI if not specified in the command.

Router#show bgp bmp table summary      
---------------------------------------------- 
Route monitoring mode : Inbound Post-Policy 
---------------------------------------------- 
Address-Family : IPv4 Unicast 
------------------------------------ 
BMP prefix table version : 12 

Number of prefixes: 6, paths: 6
The following examples displays the output when you run the show bgp bmp table command with the neighbor filter. It displays the prefixes leanrt from the specified neighbor.

Router#show bgp bmp table neighbor 23.0.2.3 

---------------------------------------------- 
Route monitoring mode : Inbound Post-Policy 
---------------------------------------------- 
Address-Family : IPv4 Unicast 
------------------------------------ 
BMP prefix table version : 11 
Number of prefixes: 5, paths: 9 

   Network            Version    Neighbor         Bit Map    Flags  Path ID 
   100.1.2.3/32       11         23.0.2.3         0x03/0xfc  0x02   1 
   100.1.3.3/32       10         23.0.2.3         0x03/0xfc  0x03   1 

 
The following example displays filtering of prefixes based on prefix version range. Only the prefixes with version within the specified range are displayed.

Router#show bgp bmp table version begin 7 end 15 
---------------------------------------------- 
Route monitoring mode : Inbound Post-Policy 
---------------------------------------------- 
Address-Family : IPv4 Unicast 
------------------------------------ 
BMP prefix table version : 11 

Number of prefixes: 5, paths: 9 
   Network            Version    Neighbor         Bit Map    Flags  Path ID 
   100.1.1.3/32       7          23.0.1.3         0x03/0xfc  0x02   1 
   100.1.2.3/32       11         23:0:2::3        0x03/0xfc  0x02   1 
    100.1.2.3/32      11         23.0.2.3         0x03/0xfc  0x02   1 
   100.1.3.3/32       10         23.0.2.3         0x03/0xfc  0x03   1 
    100.1.3.3/32      10         23.0.1.3         0x03/0xfc  0x03   1 

Processed 3 prefixes, 5 paths 

The command show bgp bmp convergence displays whether there is any pending work for BMP to perform and whether a specific address family has reached convergence. The command is applicable to Inbound Post-Policy as well as Local-RIB modes. Convergence is declared when the following conditions are met.

  • Table copy is not pending (applicable only for Inbound Post-Policy)

  • Update generation is not pending

  • All BMP servers that are not administratively shutdown are Up state

  • All BMP servers have caught up to the table version

  • Write queues of all BMP servers are empty


Router#show bgp bmp convergence  
---------------------------------------------- 
Route monitoring mode : Inbound Post-Policy 
---------------------------------------------- 
Address-Family : IPv4 Unicast 
------------------------------------ 
Converged. 

All paths from monitored neighbors are sent to the BMP servers that are not administratively shutdown.
BMP write queue is empty. 
The following example displays an address family that is converged. Since BMP mode is not specified, convergence details are shown for the default mode inbound post-policy.

Router#show bgp bmp vpnv4 unicast convergence  
---------------------------------------------- 
Route monitoring mode : Inbound Post-Policy 
---------------------------------------------- 
Address-Family : VPNv4 Unicast 
------------------------------------ 

Converged. 

All paths from monitored neighbors are sent to the BMP servers that are not administratively shutdown. 
BMP write queue is empty. 
The following example displays the convergence details for all AFIs and SAFIs. Since BMP mode is not specified, convergence information is shown for the default mode inbound post-policy.

Router#show bgp bmp all all convergence       
---------------------------------------------- 

Route monitoring mode : Inbound Post-Policy 
---------------------------------------------- 
Address-Family : VPNv4 Unicast 
------------------------------------ 
Converged. 

All paths from monitored neighbors are sent to the BMP servers that are not administratively shutdown. 

BMP write queue is empty. 

Address-Family : VPNv6 Unicast 
------------------------------------ 
Converged. 

All paths from monitored neighbors are sent to the BMP servers that are not administratively shutdown. 

BMP write queue is empty. 

Address-Family : IPv4 Labeled-unicast 
------------------------------------ 
Converged. 

All paths from monitored neighbors are sent to the BMP servers that are not administratively shutdown. 

BMP write queue is empty. 

Address-Family : IPv6 Labeled-unicast 
------------------------------------ 

Converged. 

All paths from monitored neighbors are sent to the BMP servers that are not administratively shutdown. 

BMP write queue is empty. 

Address-Family : IPv4 Unicast 
------------------------------------ 
Converged. 

All paths from monitored neighbors are sent to the BMP servers that are not administratively shutdown. 

BMP write queue is empty. 

Address-Family : IPv6 Unicast 
------------------------------------ 
Converged. 

All paths from monitored neighbors are sent to the BMP servers that are not administratively shutdown. 

BMP write queue is empty. 
The following example displays the convergence details for BMP mode inbound post-policy explicitly specified. The output of this command is same as that of show bgp bmp mode convergence as the explicitly specified mode is same as the default mode.

Router#show bgp bmp mode inbound post-policy convergence  
---------------------------------------------- 
Route monitoring mode : Inbound Post-Policy 
---------------------------------------------- 
Address-Family : IPv4 Unicast 
------------------------------------ 

Converged. 

All paths from monitored neighbors are sent to the BMP servers that are not administratively shutdown. 
BMP write queue is empty. 
The following examples displays the convergence details for BMP Local-RIB.

Router#show bgp bmp mode local-rib convergence            
---------------------------------------------- 
Route monitoring mode : Local-RIB 
---------------------------------------------- 
Address-Family : IPv4 Unicast 
------------------------------------ 

Converged. 

All paths from monitored neighbors are sent to the BMP servers that are not administratively shutdown. 
BMP write queue is empty. 
The following examples displays the convergence details for non-default address family VPNv4 Unicast.

Router#show bgp bmp mode local-rib vpnv4 unicast convergence  

---------------------------------------------- 
Route monitoring mode : Local-RIB 
---------------------------------------------- 
Address-Family : VPNv4 Unicast 
------------------------------------ 

Converged. 

All paths from monitored neighbors are sent to the BMP servers that are not administratively shutdown. 
BMP write queue is empty. 
The following examples displays the convergence details for all address families for BMP mode Local-RIB.

Router#show bgp bmp mode local-rib all all convergence 
---------------------------------------------- 
Route monitoring mode : Local-RIB 
---------------------------------------------- 
Address-Family : VPNv4 Unicast 
------------------------------------ 
Converged. 

All paths from monitored neighbors are sent to the BMP servers that are not administratively shutdown. 
BMP write queue is empty. 

Address-Family : VPNv6 Unicast 
------------------------------------ 
Converged. 

All paths from monitored neighbors are sent to the BMP servers that are not administratively shutdown. 
BMP write queue is empty. 

Address-Family : IPv4 Labeled-unicast 
------------------------------------ 
Converged. 

All paths from monitored neighbors are sent to the BMP servers that are not administratively shutdown. 

BMP write queue is empty. 

Address-Family : IPv6 Labeled-unicast 
------------------------------------ 
Converged. 

All paths from monitored neighbors are sent to the BMP servers that are not administratively shutdown. 
BMP write queue is empty. 

Address-Family : IPv4 Unicast 
------------------------------------ 
Converged. 

All paths from monitored neighbors are sent to the BMP servers that are not administratively shutdown. 
BMP write queue is empty. 

Address-Family : IPv6 Unicast 
------------------------------------ 
Converged. 

All paths from monitored neighbors are sent to the BMP servers that are not administratively shutdown. 
BMP write queue is empty. 
The following example shows the convergence details of an address family that has not converged (in this case the default address family ipv4 unicast since it is not specified in the command). In this example, in case of Inbound Post-Policy, convergence is not reached as the servers are in Down state and they have not caught up to the Table version:

Router#show bgp bmp mode all convergence 
---------------------------------------------- 
Route monitoring mode : Inbound Post-Policy 
---------------------------------------------- 
Address-Family : IPv4 Unicast 
------------------------------------ 
Not converged. 
All paths from monitored neighbors may not be sent to the BMP servers that are not administratively shutdown. 

Table copy pending: N 
Update generation pending: N 
Table version: 108 

Server  Caught-Up Version  Queue Size  State 
------------------------------------------------- 
      1                  0           0  Down 

      2                  0           0  Down 
---------------------------------------------- 
Route monitoring mode : Local-RIB 
---------------------------------------------- 
Address-Family : IPv4 Unicast 
------------------------------------ 
Not converged. 

All paths from monitored neighbors may not be sent to the BMP servers that are not administratively shutdown. 

Table copy pending: N 
Update generation pending: N 
Table version: 28 

Server  Caught-Up Version  Queue Size  State 
------------------------------------------------- 
      1                  0           0  Down 

      2                  0           0  Down 
In the following example of the show bgp bmp mode all convergence command, convergence is reached for BMP mode local-rib and address-family ipv4 unicast. However, the same is not yet reached for BMP mode inbound post-policy and address-family ipv4 unicast.

Router#show bgp bmp mode all convergence 
---------------------------------------------- 
Route monitoring mode : Inbound Post-Policy 
---------------------------------------------- 
Address-Family : IPv4 Unicast 
------------------------------------ 
Not converged. 

All paths from monitored neighbors may not be sent to the BMP servers that are not administratively shutdown. 

Table copy pending: N 
Update generation pending: Y 
Table version: 108 

 Server  Caught-Up Version  Queue Size  State 
------------------------------------------------- 
      1                  0           0  Up 

      2                  0           0  Up 
---------------------------------------------- 
Route monitoring mode : Local-RIB 
---------------------------------------------- 
Address-Family : IPv4 Unicast 
------------------------------------ 

Converged. 

All paths from monitored neighbors are sent to the BMP servers that are not administratively shutdown. 
BMP write queue is empty. 
In the following example of the same command, convergence is still not reached for BMP mode inbound post-policy. In this case the server are in Up state and they have caught up to the Table version. However, the queue for Server 1 is not empty.

Router#show bgp bmp mode all convergence 
---------------------------------------------- 
Route monitoring mode : Inbound Post-Policy 
---------------------------------------------- 
Address-Family : IPv4 Unicast 
------------------------------------ 
Not converged. 

All paths from monitored neighbors may not be sent to the BMP servers that are not administratively shutdown. 

Table copy pending: N 
Update generation pending: N 
Table version: 108 

Server  Caught-Up Version  Queue Size  State 
------------------------------------------------- 
      1                108         432  Up 

      2                108           0  Up 
---------------------------------------------- 
Route monitoring mode : Local-RIB 
---------------------------------------------- 
Address-Family : IPv4 Unicast 
------------------------------------ 

Converged. 

All paths from monitored neighbors are sent to the BMP servers that are not administratively shutdown. 
BMP write queue is empty. 

Note


The origin of prefixes conveyed in the Inbound Post-Policy view conveys is the actual BGP peers. Whereas the origin of prefixes conveyed in the Local-RIB view is the Local-RIB itself. The Local-RIB is viewed as a set of emulated peers from which the prefixes originate. Each emulated peer is associated with the prefixes belonging to one VRF (all address families inclusive) that are conveyed to one BMP server. In other words, one emulated peer is per-VRF per-BMP server. So, if BGP has n VRFs and there are m BMP servers, then n x m emulated peers are created automatically.

The internal construct of the emulated peer is largely same as that of the normal BGP peer. So, the output of some of the emulated peers related commands resembles that of the BGP peer related commands.


The following example shows the details of the Local RIB emulated peers.


Note


This keyword emulated-peers is applicable only to BMP mode Local RIB.



Router#show bgp bmp mode local-rib emulated-peers 
BGP emulated peer is 0.0.0.1 
Remote AS 2, local AS 2, internal link 
Remote router ID 100.1.1.2 
  BGP state = Established, up for 08:05:38 
  Previous State: Idle 
  Last Received Message: None 
  BGP Monitoring(BMP) activated for servers:  
      1 
  Multi-protocol capability received 
  Emulated peer capabilities: 
    Route refresh: advertised (old + new) 
    4-byte AS: advertised and received 
    Address family IPv4 Unicast: advertised and received 
    Address family IPv4 Labeled-unicast: advertised and received 
    Address family VPNv4 Unicast: advertised and received 
    Address family IPv6 Unicast: advertised and received 
    Address family IPv6 Labeled-unicast: advertised and received 
    Address family VPNv6 Unicast: advertised and received 
  Received 0 messages, 0 notifications, 0 in queue 
  Sent 20 messages, 0 notifications, 0 in queue 
  Minimum time between advertisement runs is 15 secs 

For Address Family: IPv4 Unicast 
  BMP emulated peer version 9 
  Update group: 0.9 Filter-group: 0.8  No Refresh request being processed 
  Route-Reflector Client 
    Extended Nexthop Encoding: advertised and received 
  Route refresh request: received 0, sent 0 
  Prefix advertised 4, suppressed 0, withdrawn 0 
  AIGP is enabled 
  Last ack version 9, Last synced ack version 0 
  Outstanding version objects: current 0, max 1, refresh 0 
  Additional-paths operation: None 
  Send Multicast Attributes 
  Advertise routes with local-label via Unicast SAFI 

For Address Family: IPv4 Labeled-unicast 
  BMP emulated peer version 9 
  Update group: 0.10 Filter-group: 0.9  No Refresh request being processed 
  Route-Reflector Client 
    Extended Nexthop Encoding: advertised and received 
  Route refresh request: received 0, sent 0 
  Prefix advertised 1, suppressed 0, withdrawn 0 
  AIGP is enabled 
  Last ack version 9, Last synced ack version 0 
  Outstanding version objects: current 0, max 1, refresh 0 
  Additional-paths operation: None 
  Send Multicast Attributes 

For Address Family: VPNv4 Unicast 
  BMP emulated peer version 33 
  Update group: 0.7 Filter-group: 0.6  No Refresh request being processed 
  Route-Reflector Client 
    Extended Nexthop Encoding: advertised and received 
  Route refresh request: received 0, sent 0 
  Prefix advertised 6, suppressed 0, withdrawn 0 
  AIGP is enabled 
  Last ack version 33, Last synced ack version 0 
  Outstanding version objects: current 0, max 1, refresh 0 
  Additional-paths operation: None 
  Send Multicast Attributes 

For Address Family: IPv6 Unicast 
  BMP emulated peer version 9 
  Update group: 0.11 Filter-group: 0.9  No Refresh request being processed 
  Route-Reflector Client 
    Extended Nexthop Encoding: advertised and received 
  Route refresh request: received 0, sent 0 
  Prefix advertised 4, suppressed 0, withdrawn 0 
  AIGP is enabled 
  Last ack version 9, Last synced ack version 0 
  Outstanding version objects: current 0, max 1, refresh 0 
  Additional-paths operation: None 
  Send Multicast Attributes 
  Advertise routes with local-label via Unicast SAFI 

For Address Family: IPv6 Labeled-unicast 
  BMP emulated peer version 9 
  Update group: 0.12 Filter-group: 0.10  No Refresh request being processed 
  Route-Reflector Client 
    Extended Nexthop Encoding: advertised and received 
  Route refresh request: received 0, sent 0 
  Prefix advertised 1, suppressed 0, withdrawn 0 
  AIGP is enabled 
  Last ack version 9, Last synced ack version 0 
  Outstanding version objects: current 0, max 1, refresh 0 
  Additional-paths operation: None 
  Send Multicast Attributes 

For Address Family: VPNv6 Unicast 
  BMP emulated peer version 31 
  Update group: 0.7 Filter-group: 0.6  No Refresh request being processed 
  Route-Reflector Client 
    Extended Nexthop Encoding: advertised and received 
  Route refresh request: received 0, sent 0 
  Prefix advertised 6, suppressed 0, withdrawn 0 
  AIGP is enabled 
  Last ack version 31, Last synced ack version 0 
  Outstanding version objects: current 0, max 1, refresh 0 
  Additional-paths operation: None 
  Send Multicast Attributes 

  Connections established 1; dropped 0 
  Last reset 00:00:00 
The show bgp bmp mode local-rib emulated peers command with additional keyword brief is used to display a brief summary of the Local RIB emulated peers.

Router#show bgp bmp mode local-rib emulated-peers brief 

EmuPeerId       BMPServer  VRF                   Spk    AS  OutQ  State 
0.0.0.1                 1  default                 0     2     0  Established 
0.0.0.2                 2  default                 0     2     0  Established 
0.0.0.1                 1  bar                     0     2     0  Established 
0.0.0.2                 2  bar                     0     2     0  Established 
0.0.0.1                 1  foo                     0     2     0  Established 
0.0.0.2                 2  foo                     0     2     0  Established 
The following example shows the detailed summary about Local RIB emulated peers when the detail keyword is used.

Router#show bgp bmp mode local-rib emulated-peers detail  

BGP emulated peer is 0.0.0.1 
 Remote AS 2, local AS 2, internal link 
 Remote router ID 100.1.1.2 
  BGP state = Established, up for 2d20h 
  Previous State: Idle 
  Last Received Message: None 
  BGP Monitoring(BMP) activated for servers:  
      1 
  Multi-protocol capability received 
  Emulated peer capabilities:       Adv         Rcvd 
    Route refresh:                  Yes         No 
    4-byte AS:                      Yes         Yes 
    Address family IPv4 Unicast:    Yes         Yes 
    Address family IPv4 Labeled-unicast:  Yes         Yes 
    Address family VPNv4 Unicast:   Yes         Yes 
    Address family IPv6 Unicast:    Yes         Yes 
    Address family IPv6 Labeled-unicast:  Yes         Yes 
    Address family VPNv6 Unicast:   Yes         Yes 

  Message stats: 
    OutQ depth: 0 
  Minimum time between advertisement runs is 15 secs 
The following example shows the detailed summary of a Local RIB emulated peer specified by VRF and emulated-peer address.

Router#show bgp bmp mode local-rib vrf foo emulated-peers 0.0.0.1 detail  

BGP emulated peer is 0.0.0.1, vrf foo 
 Remote AS 2, local AS 2, internal link 
 Remote router ID 100.1.1.2 
  BGP state = Established, up for 2d21h 
  Previous State: Idle 
  Last Received Message: None 
  BGP Monitoring(BMP) activated for servers:  
      1 
  Multi-protocol capability received 
  Emulated peer capabilities:       Adv         Rcvd 
    Route refresh:                  Yes         No 
    4-byte AS:                      Yes         Yes 
    Address family IPv4 Unicast:    Yes         Yes 
    Address family IPv6 Unicast:    Yes         Yes 

  Message stats: 
    OutQ depth: 0 
  Minimum time between advertisement runs is 15 secs
The following example shows the performance statistics information of a Local RIB emulated peer specified with VRF and emulated-peer address.

Router#show bgp bmp mode local-rib vrf foo emulated-peers 0.0.0.1 performance-statistics 

BGP neighbor is 0.0.0.1, vrf foo 
 Remote AS 2 
  Read 0 messages (0 bytes) in 0 calls (time spent: 0.000 secs) 
  Read partly throttled 0 times 
  Max socket read size: 0 bytes 
  Processed 0 inbound update messages (time spent: 0.000 secs) 
  Wrote 750 bytes in 2 calls (time spent: 0.000 secs) 
  Processing sub-group: wrote 6 messages in 4 calls (time spent: 0.000 secs) 
  Processing write queue: wrote 0 messages in 0 calls (time spent: 0.000 secs) 

  Received 0 messages, 0 notifications, 0 in queue 
  Sent 6 messages, 0 notifications, 0 in queue 
  NSR State: None 
  Nbr Primary fd:-1 Reset flags:0x0 Sync flags:0x0 Nbr-flags:0x3000c0 Rst retries:0 

 For Address Family: IPv4 Unicast 
  Message statistics: 
    Update messages processed: 3 
    Update messages sent: 3 
    Split-horizon Advertisement messages: sent 0, blocked 0 
    Split-horizon Withdraw messages: sent 0, blocked 0 

 For Address Family: IPv6 Unicast 
  Message statistics: 
    Update messages processed: 3 
    Update messages sent: 3 
    Split-horizon Advertisement messages: sent 0, blocked 0 
    Split-horizon Withdraw messages: sent 0, blocked 0 
The following example displays the details of all emulated peers belonging to all VRFs.

Router#show bgp bmp mode local-rib vrf all emulated-peers  

VRF: bar 
-------- 

BGP emulated peer is 0.0.0.1, vrf bar 
 Remote AS 2, local AS 2, internal link 
 Remote router ID 100.1.1.2 
  BGP state = Established, up for 2d22h 
  Previous State: Idle 
  Last Received Message: None 
  BGP Monitoring(BMP) activated for servers:  
      1 
  Multi-protocol capability received 
  Emulated peer capabilities: 
    Route refresh: advertised (old + new) 
    4-byte AS: advertised and received 
    Address family IPv4 Unicast: advertised and received 
    Address family IPv6 Unicast: advertised and received 
  Received 0 messages, 0 notifications, 0 in queue 
  Sent 6 messages, 0 notifications, 0 in queue 
  Minimum time between advertisement runs is 15 secs 

 For Address Family: IPv4 Unicast 

  BMP emulated peer version 25 
  Update group: 0.1 Filter-group: 0.3  No Refresh request being processed 
  Route-Reflector Client 
The following example displays the details of all emulated peers belong to a specified VRF.

Router#show bgp bmp mode local-rib vrf foo emulated-peers  

BGP emulated peer is 0.0.0.1, vrf foo 
 Remote AS 2, local AS 2, internal link 
 Remote router ID 100.1.1.2 
  BGP state = Established, up for 2d22h 
  Previous State: Idle 
  Last Received Message: None 
  BGP Monitoring(BMP) activated for servers:  
      1 
  Multi-protocol capability received 
  Emulated peer capabilities: 
    Route refresh: advertised (old + new) 
    4-byte AS: advertised and received 
    Address family IPv4 Unicast: advertised and received 
    Address family IPv6 Unicast: advertised and received 
  Received 0 messages, 0 notifications, 0 in queue 
  Sent 6 messages, 0 notifications, 0 in queue 
  Minimum time between advertisement runs is 15 secs 

 For Address Family: IPv4 Unicast 

  BMP emulated peer version 25 
  Update group: 0.1 Filter-group: 0.3  No Refresh request being processed 
  Route-Reflector Client 

show bgp update out

To display address-family level update generation information, use the show bgp update out command in EXEC mode.

show bgp [ vrf vrf-name ] [ afi safi ] update out [ brief | detail ]

Syntax Description

vrf vrf-name

(Optional) Displays non-default VRF.

aft

(Optional) Displays address-family identifier.

saft

(Optional) Displays subsequent address family identifier.

brief

(Optional) Displays brief information on process level update generation.

detail

(Optional) Displays detailed information on process level update generation.

Command Default

None

Command Modes

EXEC

Command History

Release Modification

Release 7.9.1

The command displays the summary of the neighbor address-family update-group, sub-group, or refresh sub-group information.

Release 4.2.0

This command was introduced.

Usage Guidelines

To use this command, you must be in a user group associated with a task group that includes appropriate task IDs. If the user group assignment is preventing you from using a command, contact your AAA administrator for assistance.

Task ID

Task ID Operation

bgp

read

Examples

This example displays sample output from the show bgp update out command:

RP/0/RSP0/CPU0:router#show bgp update out
Address-family "IPv4 Unicast"
  Update generation status: Normal
  Update OutQ:              0 bytes (0 messages)
  AF update limit:  268435456 bytes (configured 268435456 bytes)
  EBGP Sub-group update limit:   33554432 bytes (configured 33554432 bytes)
  IBGP Sub-group update limit:   33554432 bytes (configured 33554432 bytes)

  Main routing table version: 2
  RIB version: 2
  Minimum neighbor version: 2
  AF Flags: 0x00000000
  Update-groups: 1
  Sub-groups: 1 (0 throttled)
  Refresh sub-groups: 0 (0 throttled)
  Filter-groups: 1
  Neighbors: 3

  History:
    Update OutQ Hi:                   300 bytes (1 messages)
    Update OutQ Cumulative:           600 bytes (2 messages)
    Update OutQ Discarded:              0 bytes (0 messages)
    Update OutQ Cleared:                0 bytes (0 messages)
    Last discarded from OutQ:  --- (never)
    Last cleared   from OutQ:  --- (never)
    Update generation throttled 0 times, last event --- (never)
    Update generation recovered 0 times, last event --- (never)
    Update generation mem alloc failed 0 times, last event --- (never)
          
  VRF "default", Address-family "IPv4 Unicast"
    RD flags: 0x00000001
    RD Version: 2
    Table flags: 0x00000021
    RIB version: 2
    Update-groups: 1
    Sub-groups: 1 (0 throttled)
    Refresh sub-groups: 0 (0 throttled)
    Filter-groups: 1
    Neighbors: 3
          
RP/0/RSP0/CPU0:PE51_ASR-9010#
RP/0/RSP0/CPU0:PE51_ASR-9010#
RP/0/RSP0/CPU0:PE51_ASR-9010#show bgp update out filter-group 
Thu Sep 13 01:43:48.183 DST

The command shows summary of the neighbor address-family update-group, sub-group, or refresh sub-group information. It is modified to show if the peer is configured as Static slow peer or Dynamic slow peer. The command is also modified to show summary for the slow peers only.

The show bgp ipv6 unicast update out neighbor brief command displays output for all peers.

The show bgp ipv6 unicast update out neighbor slow-peers brief commnad displays the output for the slow peers only.
Router#show bgp ipv6 unicast update out neighbor slow-peers brief

Fri Nov 18 04:53:32.903 UTC

VRF "default", Address-family "IPv6 Unicast"
  Main routing table version: 1572958
  RIB version: 1572958

Legend: (S) - Slow peer static configured
        (D) - Slow peer dynamic detected

  Neighbor        FG      SG      SG-R           UG      Status OutQ       OutQ-R     Version     Ack/Ack-R

  2020:102::1     0.52    0.24    ---            0.18    Normal 0          0          1572958     1572958 (S)
  2020:103::1     0.53    0.25    ---            0.19    Normal 0          0          1572958     1572958 (S)
  2020:104::1     0.8     0.8     ---            0.20    Normal 0          0          1572958     1572958 (S)
........
  2020:11c::1     0.43    0.14    0.14:1853      0.4     Normal 0          0          1572958     1572958/0 (D)
  2020:129::1     0.3     0.3     0.3:1833       0.6     Normal 0          0          1572958     1572958/0 (D)
  2020:149::1     0.3     0.3     0.3:1849       0.6     Normal 0          0          1572958     1572678/0 (D)
  2020:156::1     0.43    0.14    0.14:1854      0.4     Normal 0          0          1572958     1572958/0 (D)
---More----

The command displays all neighbors update information. The output is enhanced to include slow peer details. The command has also been modified to show neighbors update information of only slow peer.

Router#show bgp update out neighbor slow-peers detail
Wed Jun 1 13:34:23.605 IST

VRF "default", Address-family "IPv4 Unicast"
Main routing table version: 47521
RIB version: 47521

Neighbor 192.168.0.4
Filter-group 0.1, Refresh filter-group 0.1
Sub-group 0.1, Refresh sub-group 0.1:1
Update-group 0.2

Update OutQ: 30000 bytes (50 messages)
Refresh update OutQ: 10800 bytes (18 messages)
Filter-group pending: 68 messages

Neighbor flags: 0x40310060+0x00002026
Reset 0x00003fef, Sync 0x00000000, Ver catchup 0x00000000
GSHUT 0x00000000
Neighbor AF flags: 0x00000204+0x00020030+0x00280000+0x00000018
Capability 0x00000801, GR 0x00000000

Version: 47521
Ack version: 25021 (Synced 0)
Outstanding version count: 8
Pending target version: 0 (next resume: 0)
EOR outstanding ? [No]
Refresh version: 31521
Refresh Ack version: 0
Refresh target version: 31521 (requested 31521)
Refresh pending target version: 0 (next resume: 0)
Refresh State: SLOW-RTC
Update Message Pointer: 0x7f17d4ae6988
Last Message Enqueued: 4 secs
Slow Detection State: Dynamic Detected Slow Peer
----More-----

The show bgp update out neighbor slow-peers detail command displays the output of a specific neighbor address-family for the slow peers only.

Router#show bgp vpnv4 unicast update out neighbor 10.1.114.1 detail

Tue Nov 15 18:48:24.863 UTC

VRF "default", Address-family "VPNv6 Unicast"
  Main routing table version: 51001
  RIB version: 51001

  Neighbor 10.1.114.1
    Filter-group 0.50,   Refresh filter-group 0.50
    Sub-group 0.5,   Refresh sub-group 0.5:538
    Update-group 0.2

    Update OutQ:              147400 bytes (67 messages)
    Refresh update OutQ:      341000 bytes (155 messages)
    Filter-group pending:                   222 messages

    Neighbor flags: 0x40218060+0x00000026
      Reset 0x00003fef,   Sync 0x00000000,   Ver catchup 0x00000000
      GSHUT 0x00000000,   NbrInfo 0x00000044
    Neighbor AF flags: 0x00110224+0x10060020+0x00000000+0x00000048
      Capability 0x00000001,   GR 0x00000000
      DR 0x00000000,   DR In use 0x00000000
      LLGR 0x00000000

    Version: 51001
    Ack version: 25750 (Synced 0)
    Outstanding version count: 0
    Pending target version: 0 (next resume: 0)
    EOR outstanding ?  [No]
    Refresh version: 25750
    Refresh Ack version: 0
    Refresh target version: 25750 (requested 50251)
    Refresh pending target version: 0 (next resume: 0)
    Refresh State: SLOW
    Update Message Pointer: 0x7f17d4bc6918
    Last Message Enqueued: 20 secs
    Slow Detection State: Dynamic Detected Slow Peer
----More----

Where:

  • Refresh State indicates the state of the refresh sub-group. The states can be any of the following:

    • Not-In-Refresh: Refresh sub-group is not present

    • RR: Refresh sub-group is processing refresh request update

    • SLOW: Refresh sub-group is processing slow peer update

    • RTC: Refresh sub-group is processing RTC incremental update

    • SLOW-RTC: Refresh sub-group is processing both slow peer and RTC incremental update

  • Last Message Enqueued indicates the time since the last update message was enqueued to the neighbor address-family. If this time exceeds the neighbor address-family slow peer detection threshold time, then it meets one of the conditions for detection of slow peer.

  • Slow Detection State indicates the operational type of slow peer. The type can be any of the following:

    • Static Slow Peer: Neighbor address-family is static slow peer

    • Dynamic Detected Slow Peer: Neighbor address-family is dynamic detected slow peer

    • Not slow peer: Neighbor address-family is not a slow peer

show bgp update in error process

To display process level update inbound error-handling information, use the show bgp update in error process command in EXEC mode.

show bgp update in error process [brief | detail]

Syntax Description

brief

(Optional) Displays brief information on process level update generation.

detail

(Optional) Displays detailed information on process level update generation.

Command Default

None

Command Modes

EXEC

Command History

Release

Modification

Release 4.2.0

This command was introduced.

Usage Guidelines

To use this command, you must be in a user group associated with a task group that includes appropriate task IDs. If the user group assignment is preventing you from using a command, contact your AAA administrator for assistance.

Task ID

Task ID

Operations

bgp

read

Examples

This example displays sample output from the show bgp update in error process command:

RP/0/RSP0/CPU0:router#show bgp update in error process    

Basic Update error-handling:
  EBGP: [Enabled]
  IBGP: [Enabled]
Extended Update error-handling:
  EBGP: [Disabled]
  IBGP: [Disabled]

Malformed Update messages: 0
Neighbors that received malformed Update messages: 0
Last malformed Update received: --- (never)

show bgp update out filter-group

To display update generation information at filter-group level, show bgp update out filter-group command in EXEC mode.

show bgp [vrf vrf-name] [afi safi] update out filter-group [fg-process-id] [brief | detail]

Syntax Description

vrf vrf-name

Specifies the non-default VRF.

afi safi

Specifies the address family and subsequent address family identifiers.

fg-process-id

Specifies the filter-group process ID in <x.y> format. Range is < 0-15>.<0-4294967295>.

brief

(Optional) Displays brief information on filter-group level update generation

detail

(Optional) Displays detailed information on filter-group level update generation.

Command Default

None

Command Modes

EXEC

Command History

Release

Modification

Release 4.2.0

This command was introduced

Usage Guidelines

To use this command, you must be in a user group associated with a task group that includes appropriate task IDs. If the user group assignment is preventing you from using a command, contact your AAA administrator for assistance.

Task ID

Task ID

Operations

bgp

read

Examples

This example displays sample output from show bgp update out filter-group command:

show bgp update out process

To display process level update generation information, use the show bgp update out process command in EXEC mode.

show bgp update out process [brief | detail]

Syntax Description

brief

(Optional) Displays brief information on process level update generation.

detail

(Optional) Displays detailed information on process level update generation.

Command Default

None

Command Modes

EXEC

Command History

Release Modification

Release 4.2.0

This command was introduced.

Usage Guidelines

To use this command, you must be in a user group associated with a task group that includes appropriate task IDs. If the user group assignment is preventing you from using a command, contact your AAA administrator for assistance.

Task ID

Task ID Operation

bgp

read

Examples

This example displays sample output from the show bgp update out process brief command:

RP/0/RSP0/CPU0:router#show bgp update out process 
Wed Sep 12 08:26:04.308 DST

Update generation status: Normal
Update OutQ:           0 bytes (0 messages)
Update limit:  536870912 bytes (configured 536870912 bytes)

Update generation logging:  [Disabled]

  Address-family Status    Limit      OutQ       UG   SG(Thr)    SG-R(Thr) Nbrs

  IPv4 Unicast   Normal    268435456  0          1    1(0)       0(0)       3   
  L2VPN VPLS     Normal    268435456  0          1    1(0)       0(0)       3   

History:
  Update OutQ Hi:                   300 bytes (1 messages)
  Update OutQ Cumulative:          1200 bytes (4 messages)
  Update OutQ Discarded:              0 bytes (0 messages)
  Update OutQ Cleared:                0 bytes (0 messages)
  Last discarded from OutQ:  --- (never)
  Last cleared   from OutQ:  --- (never)
  Update generation throttled 0 times, last event --- (never)
  Update generation recovered 0 times, last event --- (never)
  Update generation mem alloc failed 0 times, last event --- (never)

show bgp update out sub-group

To display sub-group update generation information, use the show bgp update out sub-group command in EXEC mode.

show bgp [vrf vrf-name] [afi safi] update out [update-group ug-index] sub-group [sg-index] [brief | detail]

Syntax Description

vrf vrf-name

(Optional) Displays non-default VRF.

aft

(Optional) Displays address-family identifier.

saft

(Optional) Displays subsequent address family identifier.

brief

(Optional) Displays brief information on process level update generation.

detail

(Optional) Displays detailed information on process level update generation.

ug-index

(Optional) Displays the update-group process ID in <x.y> format.

sg-index

(Optional) displays the sub-group process ID in <x.y> format.

Command Default

None

Command Modes

EXEC

Command History

Release Modification

Release 4.2.0

This command was introduced.

Usage Guidelines

To use this command, you must be in a user group associated with a task group that includes appropriate task IDs. If the user group assignment is preventing you from using a command, contact your AAA administrator for assistance.

Task ID

Task ID Operation

bgp

read

Examples

This example displays sample output from the show bgp update out sub-group command:


RP/0/RSP0/CPU0:router#show bgp update out sub-group 

VRF "default", Address-family "IPv4 Unicast"
  Main routing table version: 2
  RIB version: 2

  SG             UG      Status    Limit      OutQ       SG-R Nbrs Version    ()

  0.2            0.2     Normal    33554432   0          0    3    2          ()
RP/0/RSP0/CPU0:PE51_ASR-9010#

This table describes the significant fields shown in the display:

show bgp update out update-group

To display update-group update generation information, use the show bgp update out update-group command in EXEC mode.

show bgp [vrf vrf-name] [afi safi] update out update-group [ug-index] [brief | detail]

Syntax Description

vrf vrf-name

(Optional) Displays non-default VRF.

aft

(Optional) Displays address-family identifier.

saft

(Optional) Displays subsequent address family identifier.

brief

(Optional) Displays brief information on process level update generation.

detail

(Optional) Displays detailed information on process level update generation.

ug-index

(Optional) Displays the update-group process ID in <x.y> format.

Command Default

None

Command Modes

EXEC

Command History

Release Modification

Release 4.2.0

This command was introduced.

Usage Guidelines

To use this command, you must be in a user group associated with a task group that includes appropriate task IDs. If the user group assignment is preventing you from using a command, contact your AAA administrator for assistance.

Task ID

Task ID Operation

bgp

read

Examples

This example shows the significant fields on display form the show bgp update out update-group command:


RP/0/RSP0/CPU0:router#show bgp update out sub-group 

VRF "default", Address-family "IPv4 Unicast"
  Main routing table version: 2
  RIB version: 2

  SG             UG      Status    Limit      OutQ       SG-R Nbrs Version    ()

  0.2            0.2     Normal    33554432   0          0    3    2          ()
RP/0/RSP0/CPU0:PE51_ASR-9010#show bgp update ou update-group 
Wed Sep 12 08:37:24.756 DST

VRF "default", Address-family "IPv4 Unicast"

  UG      OutQ       SG(Thr)    SG-R(Thr)  FG   Nbrs

  0.2     0          1(0)       0(0)       1    3  

show bgp vrf update in error

To display VRF level update inbound error-handling information, use the show bgp vrf update in error command in EXEC mode.

show bgp [vrf vrf-name] update in error [brief | detail]

vrf vrf-name

(Optional) Displays non-default VRF.

brief

(Optional) Displays brief information.

detail

(Optional) Displays detailed information.

Command Default

None

Command Modes

EXEC

Command History

Release

Modification

Release 4.2.0

This command was introduced.

Usage Guidelines

To use this command, you must be in a user group associated with a task group that includes appropriate task IDs. If the user group assignment is preventing you from using a command, contact your AAA administrator for assistance.

Task ID

Task ID

Operations

bgp

read

Examples

This example displays sample output from show bgp vrf vrf1 update in error command:

RP/0/RSP0/CPU0:router#show bgp update in error 

VRF "default"
  Malformed Update messages: 0
  Neighbors that received malformed Update messages: 0
  Last malformed update received: --- (never)

show bgp advertised

To display advertisements for neighbors or a single neighbor, use the show bgp advertised command in EXECEXEC mode mode.

show bgp [ipv4 { all | labeled-unicast | multicast | tunnel | unicast }] advertised [neighbor ip-address] [standby] [summary]

show bgp [ ipv6 { all | labeled-unicast | multicast | unicast}] advertised [neighbor ip-address] [standby] [summary]

show bgp [ all { all | labeled-unicast | multicast | tunnel | unicast }] advertised [neighbor ip-address] [standby] [summary]

show bgp [ vpnv4 unicast [rd rd-address]] advertised [neighbor ip-address] [standby] [summary]

show bgp [ vpnv6 unicast [ rd rd-address]] advertised [neighbor ip-address] [standby] [summary]

show bgp [ vrf {vrf-name | all} [ ipv4 | {labeled-unicast | unicast} | ipv6 unicast]] advertised [neighbor ip-address] [standby] [summary]

Syntax Description

ipv4

(Optional) Specifies IP Version 4 address prefixes.

unicast

(Optional) Specifies unicast address prefixes.

multicast

(Optional) Specifies multicast address prefixes.

labeled-unicast

(Optional) Specifies labeled unicast address prefixes.

all

(Optional) For address family, specifies prefixes for all address families.

tunnel

(Optional) Specifies tunnel address prefixes.

ipv6

(Optional) Specifies IP Version 6 address prefixes.

vpnv4 unicast

(Optional) Specifies VPNv4 unicast address families.

rd rd-address

(Optional) Displays routes with a specific route distinguisher.

vrf

(Optional) Specifies VPN routing and forwarding (VRF) instance.

vrf-name

(Optional) Name of a VRF.

all

(Optional) For VRF, specifies all VRFs.

ipv4 { unicast | labeled-unicast }

(Optional) For VRF, specifies IPv4 unicast or labeled-unicast address families.

ipv6 unicast

(Optional) For VRF, specifies IPv6 unicast address families.

neighbor

(Optional) Previews advertisements for a single neighbor. If the neighbor keyword is omitted, then the advertisements for all neighbors are displayed.

ip-address

(Optional) IP address of the neighbor.

summary

(Optional) Displays a summary of advertisements.

Command Default

If no address family or subaddress family is specified, the default address family and subaddress family specified using the set default-afi and set default-safi commands are used.

Command Modes

EXECEXEC mode

Command History

Release

Modification

Release 3.7.2

This command was introduced.

Usage Guidelines

To use this command, you must be in a user group associated with a task group that includes appropriate task IDs. If the user group assignment is preventing you from using a command, contact your AAA administrator for assistance.


Note


The set default-afi command is used to specify the default address family for the session, and the set default-safi command is used to specify the default subaddress family for the session. See the System Management Command Reference for Cisco ASR 9000 Series Routers for detailed information and syntax for the set default-afi and set default-safi commands. If you do not specify a default address family, the default address family is IPv4. If you do not specify a default subaddress family, the default subaddress family is unicast.


BGP contains a separate routing table for each address family and subaddress family combination that is configured. The address family and subaddress family options specify the routing table to be examined. If the all keyword is specified for the address family or subaddress family, each matching routing table is examined in turn.

Use the show bgp advertised command to display the routes that have been advertised to peers or a specific peer. To preview advertisements that would be sent to a peer under a particular policy, even if the corresponding update messages have not been generated yet, use the show bgp policy command.


Note


When you issue the show bgp advertised command, a route is not displayed in the output unless an advertisement for that route has already been sent (and not withdrawn). If an advertisement for the route has not yet been sent, the route is not displayed.


Use the summary keyword to display a summary of the advertised routes. If you do not specify the summary keyword, the software displays detailed information about the advertised routes.


Note


The show bgp advertised command does not display the application of any outbound policy in the route details it displays. Consequently, this command provides only an indication of whether a particular route has been advertised, rather than details of which attributes were advertised. Use the show bgp policy sent-advertisements command to display the attributes that are advertised.


Task ID

Task ID

Operations

bgp

read

Examples

The following is sample output from the show bgp advertised command in EXECEXEC mode mode:


Router# show bgp advertised neighbor 10.0.101.4 summary 

Network     Next Hop            From          AS Path
1.1.1.0/24  10.0.101.1          10.0.101.1    2 3 222 333 444 555 i
1.1.2.0/24  10.0.101.1          10.0.101.1    3 4 5 6 7 i
1.1.3.0/24  10.0.101.1          10.0.101.1    77 88 33 44 55 99 99 99 i
1.1.4.0/24  10.0.101.1          10.0.101.1    2 5 6 7 8 i
1.1.7.0/24  10.0.101.1          10.0.101.1    3 5 i
1.1.8.0/24  10.0.101.1          10.0.101.1    77 88 99 99 99 i

This table describes the significant fields shown in the display.

Table 5. show bgp advertised neighbor summary Field Descriptions

Field

Description

Network

IP prefix and prefix length for a network.

Next Hop

IP address of the next system that is used when a packet is forwarded to the destination network. An entry of 0.0.0.0 indicates that the router has a non-BGP route to this network.

From

IP address of the peer that advertised this route.

AS Path

AS path of the peer that advertised this route.

Local

Indicates the route originated on the local system.

Local Aggregate

Indicates the route is an aggregate created on the local system.

Advertised to

Indicates the peer to which this entry was advertised. This field is used in the output when displaying a summary of the advertisements to all neighbors.

The following is sample output from the show bgp advertised command for detailed advertisement information:


Router# show bgp advertised neighbor 172.72.77.1

172.16.0.0/24 is advertised to 172.72.77.1
  Path info:
    neighbor: Local            neighbor router id: 172.74.84.1
    valid  redistributed  best  
  Attributes after inbound policy was applied:
 next hop: 0.0.0.0
    MET ORG AS 
    origin: incomplete  metric: 0  
    aspath:
10.52.0.0/16 is advertised to 172.72.77.1
  Path info:
    neighbor: Local Aggregate  neighbor router id: 172.74.84.1
    valid  aggregated  best  
  Attributes after inbound policy was applied:
 next hop: 0.0.0.0
    ORG AGG ATOM 
    origin: IGP  aggregator: 172.74.84.1 (1)  
    aspath:

This table describes the significant fields shown in the display.

Table 6. show bgp advertised neighbor Field Descriptions

Field

Description

is advertised to

IP address of the peer to which this route has been advertised. If the route has been advertised to multiple peers, the information is shown separately for each peer.

neighbor

IP address of the peer that advertised this route, or one of the following:

Local—Route originated on the local system.

Local Aggregate—Route is an aggregate created on the local system.

neighbor router id

BGP identifier for the peer, or the local system if the route originated on the local system.

Not advertised to any peer

Indicates the no-advertise well-known community is associated with this route. Routes with this community are not advertised to any BGP peers.

Not advertised to any EBGP peer

Indicates the no-export well-known community is associated with this route. Routes with this community are not advertised to external BGP peers, even if those external peers are part of the same confederation as the local router.

Not advertised outside the local AS

Indicates the local-AS well-known community is associated with this route. Routes with this community value are not advertised outside the local autonomous system or confederation boundary.

(Received from a RR-client)

Path was received from a route reflector client.

(received-only)

This path is not used for routing purposes. It is used to support soft reconfiguration, and records the path attributes before inbound policy was applied to a path received from a peer. A path marked “received-only” indicates that either the path was dropped by inbound policy, or the path information was modified by inbound policy and a separate copy of the modified path is used for routing.

(received & used)

Indicates that the path is used both for soft reconfiguration and routing purposes. A path marked “received and used,” implies the path information was not modified by inbound policy.

valid

Path is valid.

redistributed

Path is locally sourced through redistribution.

aggregated

Path is locally sourced through aggregation.

local

Path is locally sourced through the network command.

confed

Path was received from a confederation peer.

best

Path is selected as best.

multipath

Path is one of multiple paths selected for load-sharing purposes.

dampinfo

Indicates dampening information:

Penalty—Current penalty for this path.

Flapped—Number of times the route has flapped.

In—Time (hours:minutes:seconds) since the router noticed the first flap.

Reuse in—Time (hours:minutes:seconds) after which the path is made available. This field is displayed only if the path is currently suppressed.

Attributes after inbound policy was applied

Displays attributes associated with the received route, after any inbound policy has been applied.

AGG—Aggregator attribute is present.

AS—AS path attribute is present.

ATOM—Atomic aggregate attribute is present.

COMM—Communities attribute is present.

EXTCOMM—Extended communities attribute is present.

LOCAL—Local preference attribute is present.

MET—Multi Exit Discriminator (MED) attribute is present.

next hop—IP address of the next system used when a packet is forwarded to the destination network. An entry of 0.0.0.0 indicates that the router has a non-BGP route to this network.

ORG—Origin attribute is present.

origin

Origin of the path:

IGP—Path originated from an Interior Gateway Protocol (IGP) and was sourced by BGP using a network or aggregate-address command.

EGP—Path originated from an Exterior Gateway Protocol.

incomplete—Origin of the path is not clear. For example, a route that is redistributed into BGP from an IGP.

neighbor as

First autonomous system (AS) number in the AS path.

aggregator

Indicates that the path was received with the aggregator attribute. The autonomous system number and router-id of the system that performed the aggregation are shown.

metric

Value of the interautonomous system metric, otherwise known as the MED metric.

localpref

Local preference value. This is used to determine the preferred exit point from the local autonomous system. It is propagated throughout the local autonomous system

aspath

AS path associated with the route.

community

Community attributes associated with the path. Community values are displayed in AA:NN format, except for the following well-known communities:

Local-AS—Community with value 4294967043 or hex 0xFFFFFF03. Routes with this community value are not advertised outside the local autonomous system or confederation boundary.

no-advertise—Community with value 4294967042 or hex 0xFFFFFF02. Routes with this community value are not advertised to any BGP peers.

no-export—Community with value 4294967041 or hex 0xFFFFFF01. Routes with this community are not advertised to external BGP peers, even if those peers are in the same confederation with the local router.

Extended community

Extended community attributes associated with the path. For known extended community types, the following codes may be displayed:

RT—Route target community

SoO—Site of Origin community

LB—Link Bandwidth community

Originator

Router ID of the originating router when route reflection is used.

Cluster lists

Router ID or cluster ID of all route reflectors through which the route has passed.

show bgp aggregate-contributors

To display whether an aggregate-address has aggregate-contributors, run the show bgp aggregate-contributors command in EXEC mode.

show bgp address/mask-length aggregate-contributors

Syntax Description

aggregate-contributors

Displays contributors of an aggregate-address

Command Default

None

Command Modes

EXEC mode

Command History

Release

Modification

Release 7.5.4

This command was introduced.

Usage Guidelines

  • To use this command, you must be in a user group associated with a task group that includes appropriate task IDs. If the user group assignment is preventing you from using a command, contact your AAA administrator for assistance.

  • The command is applicable for the following Address Family Indicators (AFIs):

    • IPv4 unicast

    • IPv6 unicast

Task ID

Task ID Operation

bgp

read

Examples

The following output displays all aggregate-contributors of the BGP address 250.2.2.0/24:

Router1#show bgp 250.2.2.0/24 aggregate-contributors 
Thu Dec  1 10:15:20.393 EST
BGP router identifier 192.168.0.2, local AS number 100
BGP generic scan interval 60 secs
Non-stop routing is enabled
BGP table state: Active
Table ID: 0xe0000000   RD version: 247
BGP main routing table version 247
BGP NSR Initial initsync version 22 (Reached)
BGP NSR/ISSU Sync-Group versions 247/0
BGP scan interval 60 secs

Status codes: s suppressed, d damped, h history, * valid, > best
              i - internal, r RIB-failure, S stale, N Nexthop-discard
Origin codes: i - IGP, e - EGP, ? - incomplete
   Network            Next Hop            Metric LocPrf Weight Path
*> 250.2.2.1/32       13.0.1.1                               0 105 i
*> 250.2.2.2/32       13.0.2.1                               0 105 i
*> 250.2.2.3/32       13.0.3.1                               0 101 i
*> 250.2.2.4/32       13.0.4.1                               0 101 i
*> 250.2.2.5/32       13.0.5.1                               0 102 i
*> 250.2.2.6/32       13.0.6.1                               0 102 i
*> 250.2.2.7/32       13.0.7.1                               0 103 i
*> 250.2.2.8/32       13.0.8.1                               0 103 i
*> 250.2.2.9/32       13.0.9.1                               0 104 i
*> 250.2.2.10/32      13.0.10.1                              0 104 i
*> 250.2.2.11/32      0.0.0.0                  0         32768 ?
*> 250.2.2.12/32      0.0.0.0                  0         32768 ?
*> 250.2.2.13/32      0.0.0.0                  0         32768 ?

Processed 13 prefixes, 13 paths

show bgp af-group

To display information about Border Gateway Protocol (BGP) configuration for address family groups, use the show bgp af-group command in EXECEXEC mode mode.

show bgp af-group group-name {configuration [defaults] [nvgen] | inheritance | users}

Syntax Description

group-name

Name of the address family group to display.

configuration

(Optional) Displays the effective configuration for the af-group, including any settings that have been inherited from af-groups used by this af-group.

defaults

(Optional) Displays all configuration settings, including any default settings.

nvgen

(Optional) Displays output in the format of show running-config output.

If the defaults keyword is also specified, the output is not suitable for cutting and pasting into a configuration session.

inheritance

Displays the af-groups from which this af-group inherits configuration settings.

users

Displays the neighbors, neighbor groups, and af-groups that inherit configuration from this af-group.

Command Default

No default behavior or value

Command Modes

EXECEXEC mode

Command History

Release

Modification

Release 3.7.2

This command was introduced.

Usage Guidelines

To use this command, you must be in a user group associated with a task group that includes appropriate task IDs. If the user group assignment is preventing you from using a command, contact your AAA administrator for assistance.

Use the show bgp af-group command with the group-name configuration argument and keyword to display the effective configuration of an af-group, taking into account any configuration that may be inherited from other af-groups through the use af-group command. The source of each command is shown.

If the defaults keyword is specified, all configuration for the af-group, including default values, is shown. Default configuration is identified in the show output. Use the nvgen keyword to display configuration formatted in the style of the show running-config command. This output is suitable for cutting and pasting into configuration sessions.

Use the show bgp af-group command with the group-name inheritance argument and keyword to display the address family groups from which the specified af-group inherits configuration.

Use the show bgp af-group command with the group-name users argument and keyword to display the neighbors, neighbor groups, and af-groups that inherit configuration from the specified af-group.

Task ID

Task ID

Operations

bgp

read

Examples

The following af-group configuration is used in the examples:


  af-group group3 address-family ipv4 unicast
  remove-private-AS
  soft-reconfiguration inbound
  !
  af-group group1 address-family ipv4 unicast
  use af-group group2
  maximum-prefix 2500 75 warning-only
  default-originate
  soft-reconfiguration inbound disable
  !
  af-group group2 address-family ipv4 unicast
  use af-group group3
  send-community-ebgp
  send-extended-community-ebgp
  capability orf prefix both
  
  

The following is sample output from the show bgp af-group command with the configuration keyword in EXECEXEC mode mode. The source of each command is shown in the right column. For example, default-originate is configured directly on af-group group1 , and the remove-private-AS command is inherited from af-group group2, which in turn inherits it from af-group group3.


Router# show bgp af-group group1 configuration

af-group group1 address-family ipv4 unicast
  capability orf prefix both                [a:group2]
  default-originate                         []
  maximum-prefix 2500 75 warning-only       []
  remove-private-AS                         [a:group2 a:group3]
  send-community                            [a:group2]
  send-extended-community                   [a:group2

The following is sample output from the show bgp af-group command with the users keyword:


Router# show bgp af-group group2 users

IPv4 Unicast: a:group1

The following is sample output from the show bgp af-group command with the inheritance keyword. This example shows that the specified af-group group1 directly uses the group2 af-group, which in turn uses the group3 af-group:


RP/0/RSP0RP0/CPU0:router# show bgp af-group group1 inheritance

IPv4 Unicast: a:group2 a:group3

Table 1 describes the significant fields shown in the display.

This table describes the significant fields shown in the display.

Table 7. show bgp af-group Field Descriptions

Field

Description

[ ]

Configures the command directly on the specified address family group.

a:

Indicates the name that follows is an address family group.

n:

Indicates the name that follows is a neighbor group.

[dflt]

Indicates the setting is not explicitly configured or inherited, and the default value for the setting is used. This field may be shown when the defaults keyword is specified.

<not set>

Indicates that the configuration is disabled by default. This field may be shown when the defaults keyword is specified.

show bgp attribute-key

To display all existing attribute keys, use the show bgp attribute-key command in EXEC mode.

show bgp {ipv4 | ipv6 | all | vpnv4 unicast | vrf} attribute-key

Syntax Description

ipv4

(Optional) Specifies IP Version 4 address prefixes.

unicast

(Optional) Specifies unicast address prefixes.

multicast

(Optional) Specifies multicast address prefixes.

labeled-unicast

(Optional) Specifies labeled unicast address prefixes.

all

(Optional) For address family, specifies prefixes for all address families.

tunnel

(Optional) Specifies tunnel address prefixes.

all

(Optional) For subaddress family, specifies prefixes for all subaddress families.

ipv6

(Optional) Specifies IP Version 6 address prefixes.

vpnv4-unicast

(Optional) Specifies VPNv4 unicast address families.

vrf

(Optional) Specifies VPN routing and forwarding (VRF) instance.

vrf-name

(Optional) Name of a VRF.

all

(Optional) For VRF, specifies all VRFs.

ipv4 { unicast | labeled-unicast }

(Optional) For VRF, specifies IPv4 unicast or labeled-unicast address families.

Command Default

If no address family or subaddress family is specified, the default address family and subaddress family specified using the set default-afi and set default-safi commands are used.

Command Modes

EXEC

Command History

Release

Modification

Release 3.7.2

This command was introduced.

Usage Guidelines

To use this command, you must be in a user group associated with a task group that includes appropriate task IDs. If the user group assignment is preventing you from using a command, contact your AAA administrator for assistance.


Note


The set default-afi command is used to specify the default address family for the session, and the set default-safi command is used to specify the default subaddress family for the session. See the System Management Command Reference for Cisco ASR 9000 Series Routers for detailed information and syntax for the set default-afi and set default-safi commands. If you do not specify a default address family, the default address family is IPv4. If you do not specify a default subaddress family, the default subaddress family is unicast.


Task ID

Task ID

Operations

bgp

read

Examples

The following is sample output from the show bgp attribute-key command in EXEC mode:


RP/0/RSP0/CPU0:router# show bgp all all attribute-key

Address Family: IPv4 Unicast
============================

BGP router identifier 10.0.0.1, local AS number 1
BGP generic scan interval 60 secs
BGP main routing table version 109
BGP scan interval 60 secs
Status codes: s suppressed, d damped, h history, * valid, > best
              i - internal, S stale
Origin codes: i - IGP, e - EGP, ? - incomplete
   Network            Next Hop        AttrKey
*> 1.1.0.0/16         0.0.0.0         0x00000002
*> 10.0.0.0/16        0.0.0.0         0x00000002
*> 12.21.0.0/16       0.0.0.0         0x00000002
*> 194.3.192.1/32     10.0.101.1      0x00000009
*> 194.3.192.2/32     10.0.101.1      0x00000009
*> 194.3.192.3/32     10.0.101.1      0x00000009
*> 194.3.192.4/32     10.0.101.1      0x00000009
*> 194.3.192.5/32     10.0.101.1      0x00000009

Processed 8 prefixes, 8 paths

Address Family: IPv4 Multicast
==============================

BGP router identifier 10.0.0.1, local AS number 1
BGP generic scan interval 60 secs
BGP main routing table version 15
BGP scan interval 60 secs
Status codes: s suppressed, d damped, h history, * valid, > best
              i - internal, S stale
Origin codes: i - IGP, e - EGP, ? - incomplete
   Network            Next Hop        AttrKey
*> 194.3.193.2/32     10.0.101.1      0x00000009
*> 194.3.193.3/32     10.0.101.1      0x00000009

Processed 2 prefixes, 2 paths

Address Family: IPv6 Unicast
============================

BGP router identifier 10.0.0.1, local AS number 1
BGP generic scan interval 60 secs
BGP main routing table version 19
BGP scan interval 60 secs
Status codes: s suppressed, d damped, h history, * valid, > best
              i - internal, S stale
Origin codes: i - IGP, e - EGP, ? - incomplete
   Network            Next Hop        AttrKey
*> 2222::1111/128     2222::2         0x00000009
*> 2222::1112/128     2222::2         0x00000009

Processed 2 prefixes, 2 paths

This table describes the significant fields shown in the display.

Table 8. show bgp attribute-key Field Descriptions

Field

Description

BGP router identifier

BGP identifier for the local system.

local AS number

Autonomous system number for the local system.

BGP generic scan interval

Interval (in seconds) between scans of the BGP table by a generic scanner.

BGP main routing table version

Last version of the BGP database that was installed into the main routing table.

BGP scan interval

Interval (in seconds) between scans.

Status codes

Status of the table entry. The status is displayed as a three-character field at the beginning of each line in the table. The first character may be (in order of precedence):

S—Path is stale, indicating that a graceful restart is in progress with the peer from which the route was learned.

s—Path is more specific than a locally sourced aggregate route and has been suppressed.

*—Path is valid.

The second character may be (in order of precedence):

>—Path is the best path to use for that network.

d—Path is dampened.

h—Path is a history entry, representing a route that is currently withdrawn, but that is being maintained to preserve dampening information. Such routes should never be marked as valid.

The third character may be:

i—Path was learned by an internal BGP (iBGP) session.

Origin codes

Origin of the path. The origin code is displayed at the end of each line in the table. It can be one of the following values:

i—Path originated from an Interior Gateway Protocol (IGP) and was sourced by BGP using a network or aggregate-address command.

e—Entry originated from an Exterior Gateway Protocol (EGP).

?—Origin of the path is not clear. Usually, this is a route that is redistributed into BGP from an IGP.

Network

IP prefix and prefix length for a network.

Next Hop

IP address of the next system that is used when a packet is forwarded to the destination network. An entry of 0.0.0.0 indicates that the router has a non-BGP route to this network.

AttrKey

Key associated with the route attribute.

Processed n prefixes, n paths

Number of prefixes and number of paths processed for the table.

show bgp cidr-only

To display routes with nonnatural network masks, also known as classless interdomain routing (CIDR) routes, use the show bgp cidr-only command in EXECEXEC mode mode.

show bgp [ipv4 | vrf] cidr-only [standby]

Syntax Description

ipv4

(Optional) Specifies the IP Version 4 address family.

unicast

(Optional) Specifies the unicast address family.

multicast

(Optional) Specifies the multicast address family.

labeled-unicast

(Optional) Specifies labeled unicast address prefixes.

all

(Optional) For subaddress family, specifies all subaddress families.

tunnel

(Optional) Specifies the tunnel address family.

vrf

(Optional) Specifies VPN routing and forwarding (VRF) instance.

vrf-name

(Optional) Name of a VRF.

all

(Optional) For VRF, specifies all VRFs.

ipv4 { unicast | labeled-unicast }

(Optional) For VRF, specifies IPv4 unicast or labeled-unicast address families.

Command Default

If no address family or subaddress family is specified, the default address family and subaddress family specified using the set default-afi and set default-safi commands are used. This command is applicable only for IPv4 prefixes. If the default address family is not IPv4, then the ipv4 keyword must be used.

Command Modes

EXECEXEC mode

Command History

Release

Modification

Release 3.7.2

This command was introduced.

Usage Guidelines

To use this command, you must be in a user group associated with a task group that includes appropriate task IDs. If the user group assignment is preventing you from using a command, contact your AAA administrator for assistance.


Note


The set default-afi command is used to specify the default address family for the session, and the set default-safi command is used to specify the default subaddress family for the session. See the System Management Command Reference for Cisco ASR 9000 Series Routers for detailed information and syntax for the set default-afi and set default-safi commands. If you do not specify a default address family, the default address family is IPv4. If you do not specify a default subaddress family, the default subaddress family is unicast.


Border Gateway Protocol (BGP) contains a separate routing table for each address family and subaddress family combination that has been configured. The address family and subaddress family options specify the routing table to be examined. If the all keyword is specified for subaddress family, all subaddress family routing tables are examined.

The show bgp cidr-only command applies only for IPv4 prefixes. If the ipv4 keyword is not specified and the default address family is not IPv4, the command is not available.

Use the show bgp cidr-only command to display CIDR routes. Routes that have their correct class (class A, B, or C) prefix length are not displayed.

Task ID

Task ID

Operations

bgp

read

Examples

The following is sample output from the show bgp cidr-only command in EXECEXEC mode mode:


Router# show bgp cidr-only

BGP router identifier 172.20.1.1, local AS number 1820
BGP main routing table version 2589
Dampening enabled
BGP scan interval 60 secs
Status codes: s suppressed, d damped, h history, * valid, > best
              i - internal, S stale
Origin codes: i - IGP, e - EGP, ? - incomplete
Network             Next Hop          Metric   LocPrf   Weight   Path
*> 192.0.0.0/8      192.168.72.24     0        1878     ?
*> 192.168.0.0/16   192.168.72.30     0        108      ?

This table describes the significant fields shown in the display.

Table 9. show bgp cidr-only Field Descriptions

Field

Description

BGP router identifier

BGP identifier for the local system.

local AS number

Autonomous system number for the local system.

BGP main routing table version

Last version of the BGP database that was installed into the main routing table.

Dampening enabled

Displayed if dampening is enabled for the routes in this BGP routing table.

BGP scan interval

Interval (in seconds) between scans of the BGP table specified by the address family and subaddress family.

Status codes

Status of the table entry. The status is displayed as a three-character field at the beginning of each line in the table. The first character may be (in order of precedence):

S—Path is stale, indicating that a graceful restart is in progress with the peer from which the route was learned.

s—Path is more specific than a locally sourced aggregate route and has been suppressed.

*—Path is valid.

The second character may be (in order of precedence):

>—Path is the best path to use for that network.

d—Path is dampened.

h—Path is a history entry, representing a route that is currently withdrawn, but that is being maintained to preserve dampening information. Such routes should never be marked as valid.

The third character may be:

i—Path was learned by an internal BGP (iBGP) session.

Origin codes

Origin of the path. The origin code is displayed at the end of each line in the table. It can be one of the following values:

i—Path originated from an Interior Gateway Protocol (IGP) and was sourced by BGP using a network or aggregate-address command.

e—Entry originated from an Exterior Gateway Protocol (EGP).

?—Origin of the path is not clear. Usually, this is a route that is redistributed into BGP from an IGP.

Network

IP prefix and prefix length for a network.

Next Hop

IP address of the next system that is used when a packet is forwarded to the destination network. An entry of 0.0.0.0 indicates that the router has a non-BGP route to this network.

Metric

Value of the interautonomous system metric, otherwise known as the Multi Exit Discriminator (MED) metric.

LocPrf

Local preference value. This is used to determine the preferred exit point from the local autonomous system. It is propagated throughout the local autonomous system.

Weight

Path weight. Weight is used in choosing the preferred path to a route. It is not advertised to any neighbor.

Path

Autonomous system path to the destination network. At the end of the path is the origin code for the path.

show bgp community

To display routes that have the specified Border Gateway Protocol (BGP) communities, use the show bgp community command in EXECEXEC mode mode.

show bgp [ipv4 {unicast | multicast | labeled-unicast | all | tunnel}] community community-list [exact-match]

show bgp [ipv6 {unicast}] community community-list [exact-match]

show bgp [all {unicast | multicast | labeled-unicast | all | tunnel}] community community-list [exact-match]

show bgp [vpnv4 unicast [rd rd-address]] community community-list [exact-match]

show bgp [vrf {vrf-name | all} [ipv4 | {unicast | labeled-unicast} | ipv6 unicast]] community community-list [exact-match]

Syntax Description

ipv4

(Optional) Specifies IP Version 4 address prefixes.

unicast

(Optional) Specifies unicast address prefixes.

multicast

(Optional) Specifies multicast address prefixes.

labeled-unicast

(Optional) Specifies labeled unicast address prefixes.

all

(Optional) For subaddress families, specifies prefixes for all subaddress families.

tunnel

(Optional) Specifies tunnel address prefixes.

ipv6

(Optional) Specifies IP Version 6 address prefixes.

all

(Optional) For address family, specifies prefixes for all address families.

vpnv4 unicast

(Optional) Specifies VPNv4 unicast address families.

rd rd-address

(Optional) Displays routes with a specific route distinguisher.

vrf

(Optional) Specifies VPN routing and forwarding (VRF) instance.

vrf-name

(Optional) Name of a VRF.

all

(Optional) For VRF, specifies all VRFs.

ipv4 { unicast | labeled-unicast }

(Optional) For VRF, specifies IPv4 unicast or labeled-unicast address families.

ipv6 unicast

(Optional) For VRF, specifies IPv6 unicast address families.

community

Specifies that only routes with communities specified by community-list is displayed.

community-list

Between one and seven communities. Each community can be a number in the range from 1 to 4294967295, a community specified in AA:NN format, or one of the following well-known communities:

graceful-shutdown — Reduced preference for shutdown (well-known community)

local-AS —Well-known community with value 4294967043 or hex 0xFFFFFF03. Routes with this community value are not advertised outside the local autonomous system or confederation boundary.

no-advertise —Well-known community with value 4294967042 or hex 0xFFFFFF02. Routes with this community value are not advertised to any BGP peers.

no-export —Well-known community with value4294967041 or hex 0xFFFFFF01. Routes with this community are not advertised to external BGP peers, even if those peers are in the same confederation as the local router.

internet —Well-known community whose value is not defined in BGP RFC. IOS XR BGP uses a value of 0 for the internet community. Routes with this community are advertised to all peers without any restrictions.

For the AA:NN format:

AA—Range is 0 to 65535.

NN—Range is 1 to 4294967295.

Up to seven community numbers can be specified.

exact-match

(Optional) Displays those routes that have communities exactly matching the specified communities.

Command Default

If no address family or subaddress family is specified, the default address family and subaddress family specified using the set default-afi and set default-safi commands are used.

Command Modes

EXECEXEC mode

Command History

Release

Modification

Release 3.7.2

This command was introduced.

Release 5.3.2

The graceful-shutdown keyword was added.

Usage Guidelines

To use this command, you must be in a user group associated with a task group that includes appropriate task IDs. If the user group assignment is preventing you from using a command, contact your AAA administrator for assistance.


Note


The set default-afi command is used to specify the default address family for the session, and the set default-safi command is used to specify the default subaddress family for the session. See the System Management Command Reference for Cisco ASR 9000 Series Routers for detailed information and syntax for the set default-afi and set default-safi commands. If you do not specify a default address family, the default address family is IPv4. If you do not specify a default subaddress family, the default subaddress family is unicast.


BGP contains a separate routing table for each configured address family and subaddress family combination. The address family and subaddress family options specify the routing table to be examined. If the all keyword is specified for the address family or the subaddress family, each matching routing table is examined in turn.

If more than seven communities are required, it is necessary to configure a route policy and use the show bgp route-policy command.

Use the exact-match keyword to display only those routes with a set of communities exactly matching the list of specified communities. If you omit the exact-match keyword, those routes containing at least the specified communities are displayed.

Task ID

Task ID

Operations

bgp

read

Examples

The following is sample output from the show bgp community graceful-shutdown command displaying the graceful maintenance feature information:
RP/0/0/CPU0:R4#show bgp community graceful-shutdown 
Tue Jan 27 13:36:25.006 PST
BGP router identifier 192.168.0.4, local AS number 4
BGP generic scan interval 60 secs
BGP table state: Active
Table ID: 0xe0000000   RD version: 18
BGP main routing table version 18
BGP scan interval 60 secs

Status codes: s suppressed, d damped, h history, * valid, > best
              i - internal, r RIB-failure, S stale, N Nexthop-discard
Origin codes: i - IGP, e - EGP, ? - incomplete
   Network            Next Hop            Metric LocPrf Weight Path
*  5.5.5.5/32         10.10.10.1                     88      0 1 ?

Processed 1 prefixes, 1 paths

Examples

The following is sample output from the show bgp community command in EXECEXEC mode mode:


Router# show bgp community 1820:1 exact-match

BGP router identifier 172.20.1.1, local AS number 1820
BGP main routing table version 55
Dampening enabled
BGP scan interval 60 secs
Status codes: s suppressed, d damped, h history, * valid, > best
              i - internal, S stale
Origin codes: i - IGP, e - EGP, ? - incomplete
   Network            Next Hop            Metric LocPrf Weight Path
*  10.13.0.0/16       192.168.40.24                          0 1878 704 701 200 ?
*  10.16.0.0/16       192.168.40.24                          0 1878 704 701 i

This table describes the significant fields shown in the display.

Table 10. show bgp community Field Descriptions

Field

Description

BGP router identifier

BGP identifier for the local system.

local AS number

Autonomous system number for the local system.

BGP main routing table version

Last version of the BGP database that was installed into the main routing table.

Dampening enabled

Displayed if dampening is enabled for the routes in this BGP routing table.

BGP scan interval

Interval (in seconds) between scans of the BGP table specified by the address family and subaddress family.

Status codes

Status of the table entry. The status is displayed as a three character field at the beginning of each line in the table. The first character may be (in order of precedence):

S—Path is stale, indicating that a graceful restart is in progress with the peer from which the route was learned.

s—Path is more specific than a locally sourced aggregate route and has been suppressed.

*—Path is valid.

The second character may be (in order of precedence):

>—Path is the best path to use for that network.

d—Path is dampened.

h—Path is a history entry, representing a route that is currently withdrawn, but that is being maintained to preserve dampening information. Such routes should never be marked as valid.

The third character may be:

i—Path was learned by an internal BGP (iBGP) session.

Origin codes

Origin of the path. The origin code is displayed at the end of each line in the table. It can be one of the following values:

i—Path originated from an Interior Gateway Protocol (IGP) and was advertised with a network or aggregate-address command.

e—Path originated from an Exterior Gateway Protocol (EGP).

?—Origin of the path is not clear. Usually, this is a route that is redistributed into BGP from an IGP.

Network

IP prefix and prefix length for a network.

Next Hop

IP address of the next system that is used when a packet is forwarded to the destination network. An entry of 0.0.0.0 indicates that the router has a non-BGP route to this network.

Metric

Value of the interautonomous system metric, otherwise known as the Multi Exit Discriminator (MED) metric.

LocPrf

Local preference value. This is used to determine the preferred exit point from the local autonomous system. It is propagated throughout the local autonomous system.

Weight

Path weight. Weight is used in choosing the preferred path to a route. It is not advertised to any neighbor.

Path

Autonomous system path to the destination network. At the end of the path is the origin code for the path.

show bgp convergence

To display whether a specific address family has reached convergence, use the show bgp convergence command in EXEC mode.

show bgp [ipv4 {unicast | multicast | labeled-unicast | all | tunnel}] convergence

show bgp [ipv6 {unicast}] convergence

show bgp [all {unicast | multicast | labeled-unicast | all | tunnel}] convergence

show bgp [vpnv4 unicast ] convergence

Syntax Description

ipv4

(Optional) Specifies the IP Version 4 address family.

unicast

(Optional) Specifies the unicast address family.

multicast

(Optional) Specifies the multicast address family.

labeled-unicast

(Optional) Specifies unicast address prefixes.

all

(Optional) For subaddress family, specifies all subaddress families.

tunnel

(Optional) Specifies tunnel address prefixes.

ipv6

(Optional) Specifies the IP Version 6 address family.

all

(Optional) For address family, specifies all address families.

vpnv4 unicast

(Optional) Specifies VPNv4 unicast address families.

Command Default

If no address family or subaddress family is specified, the default address family and subaddress family specified using the set default-afi and set default-safi commands are used.

Command Modes

EXEC

Command History

Release

Modification

Release 3.7.2

This command was introduced.

Usage Guidelines

To use this command, you must be in a user group associated with a task group that includes appropriate task IDs. If the user group assignment is preventing you from using a command, contact your AAA administrator for assistance.


Note


The set default-afi command is used to specify the default address family for the session, and the set default-safi command is used to specify the default subaddress family for the session. See the System Management Command Reference for Cisco ASR 9000 Series Routers for detailed information and syntax for the set default-afi and set default-safi commands. If you do not specify a default address family, the default address family is IPv4. If you do not specify a default subaddress family, the default subaddress family is unicast.


Border Gateway Protocol (BGP) contains a separate routing table for each configured address family and subaddress family combination. The address family and subaddress family options specify the routing table to be examined. If the all keyword is specified for the address family or subaddress family, each matching routing table is examined in turn.

Use the show bgp convergence command to see if there is any pending work for BGP to perform. The software checks the following conditions to determine whether the specified address family has converged. If all the conditions are true, the address family is considered converged.

  • All received updates have been processed and best routes selected.
  • All selected routes have been installed in the global Routing Information Base (RIB).
  • All selected routes have been advertised to peers, including any peers that are not established (unless those peers have been administratively shut down). See the shutdown (BGP) command for more information about administrative shutdown.

While testing that all selected routes have been advertised to peers, the show bgp convergence command checks the size of the write queue for each neighbor. Because this queue is shared by all address families, there is a small possibility that the command indicates the address family has not converged when, in fact, it has converged. This could happen if the neighbor write queue contained messages from some other address family.

If the specified address family has not converged, the show bgp convergence command output does not indicate the amount of work that is pending. To display this information, use theshow bgp summary command.

Task ID

Task ID

Operations

bgp

read

Examples

The following shows the result of using the show bgp convergence command for an address family that has converged:


RP/0/RSP0/CPU0:router# show bgp convergence

Converged.
All received routes in RIB, all neighbors updated.
All neighbors have empty write queues.

The following shows the result of using the show bgp convergence command for an address family that has not converged:


RP/0/RSP0/CPU0:router# show bgp convergence

Not converged.
Received routes may not be entered in RIB.
One or more neighbors may need updating.

This table describes the significant fields shown in the display.

Table 11. show bgp convergence Field Descriptions

Field

Description

Converged/Not converged

Specifies whether or not all routes have been installed in the RIB and updates have been generated and sent to all neighbors.

[All] Received routes...

For convergence, all routes must have been installed into the RIB and all updates must have been generated. For non-convergence, some routes may not be installed in the RIB, or some routes that have been withdrawn have not yet been removed from the RIB, or some routes that are up to date in the RIB have not been advertised to all neighbors.

[All | One or more] neighbors...

Specifies the status of neighbor updating.

show bgp dampened-paths

To display Border Gateway Protocol (BGP) dampened routes, use the show bgp dampened-paths command in EXECEXEC mode mode.

show bgp [ipv4 {unicast | multicast | labeled-unicast | all}] dampened-paths

show bgp [ipv6 {unicast}] dampened-paths

show bgp [all {unicast | multicast | labeled-unicast | all | tunnel}] dampened-paths

show bgp [vpnv4 unicast [rd rd-address]] dampened-paths

show bgp [vrf {vrf-name | all} [ipv4 | {unicast | labeled-unicast} | ipv6 unicast]] dampened-paths

Syntax Description

ipv4

(Optional) Specifies IP Version 4 address prefixes.

unicast

(Optional) Specifies unicast address prefixes.

multicast

(Optional) Specifies multicast address prefixes.

labeled-unicast

(Optional) Specifies labeled unicast address prefixes.

all

(Optional) For subaddress families, specifies prefixes for all subaddress families.

ipv6

(Optional) Specifies IP Version 6 address prefixes.

all

(Optional) For address family, specifies prefixes for all address families.

vpnv4 unicast

(Optional) Specifies VPNv4 unicast address families.

rd rd-address

(Optional) Displays routes with a specific route distinguisher.

vrf

(Optional) Specifies VPN routing and forwarding (VRF) instance.

vrf-name

(Optional) Name of a VRF.

all

(Optional) For VRF, specifies all VRFs.

ipv4 {unicast | labeled-unicast }

(Optional) For VRF, specifies IPv4 unicast or labeled-unicast address families.

ipv6 unicast

(Optional) For VRF, specifies IPv6 unicast address families.

Command Default

If no address family or subaddress family is specified, the default address family and subaddress family specified using the set default-afi and set default-safi commands are used.

Command Modes

EXECEXEC mode

Command History

Release

Modification

Release 3.7.2

This command was introduced.

Usage Guidelines

To use this command, you must be in a user group associated with a task group that includes appropriate task IDs. If the user group assignment is preventing you from using a command, contact your AAA administrator for assistance.


Note


The set default-afi command is used to specify the default address family for the session, and the set default-safi command is used to specify the default subaddress family for the session. See the System Management Command Reference for Cisco ASR 9000 Series Routers for detailed information and syntax for the set default-afi and set default-safi commands. If you do not specify a default address family, the default address family is IPv4. If you do not specify a default subaddress family, the default subaddress family is unicast.


BGP contains a separate routing table for each configured address family and subaddress family combination. The address family and subaddress family options specify the routing table to be examined. If the all keyword is specified for the address family or for the subaddress family, each matching routing table is examined in turn.

Task ID

Task ID

Operations

bgp

read

Examples

The following is sample output from the show bgp dampened-paths command in EXECEXEC mode mode:


Router# show bgp dampened-paths

BGP router identifier 10.2.0.1, local AS number 3
BGP generic scan interval 60 secs
BGP main routing table version 7
Dampening enabled
BGP scan interval 60 secs
Status codes:s suppressed, d damped, h history, * valid, > best
              i - internal, S stale

Origin codes:i - IGP, e - EGP, ? - incomplete
   Network            From            Reuse    Path
*d 10.0.0.0           10.0.101.35     00:01:20 35 i

This table describes the significant fields shown in the display.

Table 12. show bgp dampened-paths Field Descriptions

Field

Description

BGP router identifier

BGP identifier for the local system.

local AS number

Autonomous system number for the local system.

BGP generic scan interval

Interval (in seconds) between scans of the BGP table by a generic scanner.

BGP main routing table version

Last version of the BGP database that was installed into the main routing table.

Dampening enabled

Displayed if dampening is enabled for the routes in this BGP routing table.

BGP scan interval

Interval (in seconds) between scans of the BGP table specified by the address family and subaddress family.

Status codes

Status of the table entry. The status is displayed as a three-character field at the beginning of each line in the table. The first character may be (in order of precedence):

S—Path is stale, indicating that a graceful restart is in progress with the peer from which the route was learned.

s—Path is more specific than a locally sourced aggregate route and has been suppressed.

*—Path is valid.

The second character may be (in order of precedence):

>—Path is the best path to use for that network.

d—Path is dampened.

h—Path is a history entry, representing a route that is currently withdrawn, but that is being maintained to preserve dampening information. Such routes should never be marked as valid.

The third character may be:

i—Path was learned by an internal BGP (iBGP) session.

Origin codes

Origin of the path. The origin code is displayed at the end of each line in the table. It can be one of the following values:

i—Path originated from an Interior Gateway Protocol (IGP) and was advertised with a network or aggregate-address command.

e—Path originated from an Exterior Gateway Protocol (EGP).

?—Origin of the path is not clear. Usually, this is a route that is redistributed into BGP from an IGP.

Network

IP prefix and prefix length for a network.

From

Neighbor from which the route was received.

Reuse

Time (in hours:minutes:seconds) after which the path is made available.

Path

Autonomous system path to the destination network. At the end of the path is the origin code for the path.

show bgp flap-statistics

To display information about Border Gateway Protocol (BGP) paths that have flapped, use the show bgp flap-statistics command in EXECEXEC mode mode.

show bgp [ipv4 {unicast | multicast | labeled-unicast | all}] flap-statistics [regexp regular-expression| route-policy route-policy-name | cidr-only | {ip-address | {mask | /prefix-length}}] [longer-prefixes] [detail]

show bgp [ipv6 {unicast}] flap-statistics [regexp regular-expression| route-policy route-policy-name | cidr-only | {ip-address | {mask | /prefix-length}}] [longer-prefixes] [detail]

show bgp [all {unicast | multicast | labeled-unicast | all}] flap-statistics [regexp regular-expression| route-policy route-policy-name | cidr-only | {ip-address | {mask | /prefix-length}}] [longer-prefixes] [detail]

show bgp [vpnv4 unicast [rd rd-address]] flap-statistics [regexp regular-expression| route-policy route-policy-name | cidr-only | {ip-address | {mask | /prefix-length}}] [longer-prefixes] [detail]

show bgp [vrf {vrf-name | all} [ipv4 | {unicast | labeled-unicast} | ipv6 unicast]] flap-statistics [regexp regular-expression| route-policy route-policy-name | cidr-only | {ip-address | {mask | /prefix-length}}] [longer-prefixes] [detail]

Syntax Description

ipv4

(Optional) Specifies IP Version 4 address prefixes.

unicast

(Optional) Specifies unicast address prefixes.

multicast

(Optional) Specifies multicast address prefixes.

labeled-unicast

(Optional) Specifies labeled unicast address prefixes.

all

(Optional) For subaddress families, specifies prefixes for all subaddress families.

ipv6

(Optional) Specifies IP Version 6 address prefixes.

all

(Optional) For address family, specifies prefixes for all address families.

vpnv4 unicast

(Optional) Specifies VPNv4 unicast address families.

rd rd-address

(Optional) Displays routes with a specific route distinguisher.

vrf

(Optional) Specifies VPN routing and forwarding (VRF) instance.

vrf-name

(Optional) Name of a VRF.

all

(Optional) For VRF, specifies all VRFs.

ipv4 { unicast | labeled-unicast }

(Optional) For VRF, specifies IPv4 unicast or labeled-unicast address families.

ipv6 unicast

(Optional) For VRF, specifies IPv6 unicast address families.

regexp regular-expression

(Optional) Displays flap statistics for all paths that match the regular expression.

route-policy route-policy-name

(Optional) Displays flap statistics for a route policy.

cidr-only

(Optional) Displays only routes whose prefix length does not match the classful prefix length for that network. The cidr-only keyword can be specified only if the address family is IPv4.

ip-address

(Optional) Flap statistics for a network address only.

mask

(Optional) Network mask applied to the ip-address argument.

/ prefix-length

(Optional) Length of the IP address prefix. A decimal value that indicates how many of the high-order contiguous bits of the address compose the prefix (the network portion of the address). A slash (/) must precede the decimal value.

longer-prefixes

(Optional) Displays flap statistics for the specified prefix and more-specific prefixes. The longer-prefixes keyword is available when the ip-address and mask or /prefix-length arguments are specified.

detail

(Optional) Displays dampening parameters for the path. The detail keyword cannot be specified if the longer-prefixes keyword is specified. The detail keyword is available when the ip-address argument or ip-address and mask or /prefix-length arguments are specified.

Command Default

If no address family or subaddress family is specified, the default address family and subaddress family specified using the set default-afi and set default-safi commands are used.

Command Modes

EXECEXEC mode

Command History

Release

Modification

Release 3.7.2

This command was introduced.

Usage Guidelines

To use this command, you must be in a user group associated with a task group that includes appropriate task IDs. If the user group assignment is preventing you from using a command, contact your AAA administrator for assistance.


Note


The set default-afi command is used to specify the default address family for the session, and the set default-safi command is used to specify the default subaddress family for the session. See the System Management Command Reference for Cisco ASR 9000 Series Routers for detailed information and syntax for the set default-afi and set default-safi commands. If you do not specify a default address family, the default address family is IPv4. If you do not specify a default subaddress family, the default subaddress family is unicast.


BGP contains a separate routing table for each configured address family and subaddress family combination. The address family and subaddress family options specify the routing table to be examined. If the all keyword is specified for the address family or subaddress family, each matching routing table is examined in turn.

Flap statistics are maintained only for paths if dampening is enabled using the bgp dampening command. If dampening is not enabled, the show bgp flap-statistics command does not display any paths.

If no arguments or keywords are specified, the software displays flap statistics for all paths for the specified address family. You can use the regexp , filter-list , cidr-only , and longer-prefixes options to limit the set of paths displayed.

If you specify a network address without a mask or prefix length, the longest matching prefix for the network address is displayed. When displaying flap statistics for a single route, use the detail keyword to display dampening parameters for the route.

Task ID

Task ID

Operations

bgp

read

Examples

The following is sample output from the show bgp flap-statistics command:


Router# show bgp flap-statistics

BGP router identifier 172.20.1.1, local AS number 1820
BGP main routing table version 26180
Dampening enabled
BGP scan interval 60 secs
Status codes: s suppressed, d damped, h history, * valid, > best
              i - internal, S stale
Origin codes: i - IGP, e - EGP, ? - incomplete
Network             From           Flaps Duration Reuse    Path
*d 10.0.0.0         172.20.16.177  4     00:13:31 00:18:10 100
*d 10.10.0.0        172.20.16.177  4     00:02:45 00:28:20 100

The following is sample output from the show bgp flap-statistics command with the detail keyword in EXECEXEC mode mode:


Router# show bgp flap-statistics 172.31.12.166 detail 

BGP router identifier 10.0.0.5, local AS number 1
BGP main routing table version 738
Dampening enabled
BGP scan interval 60 secs
Status codes: s suppressed, d damped, h history, * valid, > best
              i - internal, S stale
Origin codes: i - IGP, e - EGP, ? - incomplete
   Network            From            Flaps Duration Reuse    Path
 h 172.31.12.166            10.0.101.1      6     00:03:28          2 2000 3000 

   Half life        Suppress        Reuse penalty    Max. supp. time
   00:15:00         2000            750              01:00:00

This table describes the significant fields shown in the display.

Table 13. show bgp flap-statistics Field Descriptions

Field

Description

BGP route identifier

BGP identifier for the local system.

local AS number

Autonomous system number for the local system.

BGP main routing table version

Last version of the BGP database that was installed into the main routing table.

Dampening enabled

Displayed if dampening has been enabled for the routes in this BGP routing table.

BGP scan interval

Interval (in seconds) between scans of the BGP table specified by the address family and subaddress family.

Status codes

Status of the table entry. The status is displayed as a three-character field at the beginning of each line in the table. The first character may be (in order of precedence):

S—Path is stale, indicating that a graceful restart is in progress with the peer from which the route was learned.

s—Path is more specific than a locally sourced aggregate route and has been suppressed.

*—Path is valid.

The second character may be (in order of precedence):

>—Path is the best path to use for that network.

d—Path is dampened.

h—Path is a history entry, representing a route that is currently withdrawn, but that is being maintained to preserve dampening information. Such routes should never be marked as valid.

The third character may be:

i—Path was learned by an internal BGP (iBGP) session.

Origin codes

Origin of the path. The origin code is displayed at the end of each line in the table. It can be one of the following values:

i—Path originated from an Interior Gateway Protocol (IGP) and was advertised with a network or aggregate-address command.

e—Path originated from an Exterior Gateway Protocol (EGP).

?—Origin of the path is not clear. Usually, this is a route that is redistributed into BGP from an IGP.

Network

IP prefix and prefix length for a network that is dampened.

From

IP address of the peer that advertised this route.

Flaps

Number of times the route has flapped.

Duration

Time (in hours:minutes:seconds) since the first flap.

Reuse

Time (in hours:minutes:seconds) after which the path is made available.

Path

Autonomous system path of the route that is being dampened.

Half life

Half-life value used when dampening this route. The half-life is the amount of time that must elapse to reduce the reuse penalty by half. The half-life value is specified using the bgp dampening command.

Suppress

Suppress value used to dampen this route. The suppress value is the value that the penalty must exceed for the route to be suppressed. The suppress value can be configured using the bgp dampening command.

Reuse penalty

Reuse penalty used to dampen this route. The penalty must fall below the reuse penalty for the route to be unsuppressed. The reuse penalty can be configured using the bgp dampening command.

Max supp. time

Maximum length of time that the route may be suppressed due to dampening. The maximum suppress time can be configured using the bgp dampening command.

show bgp inconsistent-as

To display Border Gateway Protocol (BGP) routes originated from more than one autonomous system, use the show bgp inconsistent-as command in EXECEXEC mode mode.

show bgp [ipv4 {unicast | multicast | labeled-unicast | all | tunnel}] inconsistent-as [standby]

show bgp [ipv6 {unicast}] inconsistent-as [standby]

show bgp [all {unicast | multicast | labeled-unicast | all | tunnel}] inconsistent-as [standby]

show bgp vpnv4 unicast [rd rd-address] inconsistent-as [standby]

show bgp [vrf {vrf-name | all} [ipv4 | {unicast | labeled-unicast} | ipv6 unicast]] inconsistent-as [standby]

Syntax Description

ipv4

(Optional) Specifies IP Version 4 address prefixes.

unicast

(Optional) Specifies unicast address prefixes.

multicast

(Optional) Specifies multicast address prefixes.

labeled-unicast

(Optional) Specifies labeled unicast address prefixes.

all

(Optional) For subaddress families, specifies prefixes for all subaddress families.

tunnel

(Optional) Specifies tunnel address prefixes.

ipv6

(Optional) Specifies IP Version 6 address prefixes.

all

(Optional) For address family, specifies prefixes for all address families.

vpnv4 unicast

(Optional) Specifies VPNv4 unicast address families.

rd rd-address

(Optional) Displays routes with a specific route distinguisher.

vrf

(Optional) Specifies VPN routing and forwarding (VRF) instance.

vrf-name

(Optional) Name of a VRF.

all

(Optional) For VRF, specifies all VRFs.

ipv4 { unicast | labeled-unicast }

(Optional) For VRF, specifies IPv4 unicast or labeled-unicast address families.

ipv6 unicast

(Optional) For VRF, specifies IPv6 unicast address families.

Command Default

If no address family or subaddress family is specified, the default address family and subaddress family specified using the set default-afi and set default-safi commands are used.

Command Modes

EXECEXEC mode

Command History

Release

Modification

Release 3.7.2

This command was introduced.

Usage Guidelines

To use this command, you must be in a user group associated with a task group that includes appropriate task IDs. If the user group assignment is preventing you from using a command, contact your AAA administrator for assistance.


Note


The set default-afi command is used to specify the default address family for the session, and the set default-safi command is used to specify the default subaddress family for the session. See the System Management Command Reference for Cisco ASR 9000 Series Routers for detailed information and syntax for the set default-afi and set default-safi commands. If you do not specify a default address family, the default address family is IPv4. If you do not specify a default subaddress family, the default subaddress family is unicast.


BGP contains a separate routing table for each configured address family and subaddress family combination. The address family and subaddress family options specify the routing table to be examined. If the all keyword is specified for the address family or for the subaddress family, each matching routing table is examined in turn.

Use the show bgp inconsistent-as command to search through all prefixes in the specified BGP routing table and display the paths for any prefix that has inconsistent originating autonomous system numbers. The originating autonomous system is the last autonomous system number displayed in the path field and should be the same for all paths.

If a prefix has one or more paths originating from different autonomous systems, all paths for that prefix are displayed.

Task ID

Task ID

Operations

bgp

read

Examples

The following is sample output from the show bgp inconsistent-as command in EXECEXEC mode mode:


Router# show bgp inconsistent-as

BGP router identifier 172.20.1.1, local AS number 1820
BGP main routing table version 1129
BGP scan interval 60 secs
Status codes: s suppressed, d damped, h history, * valid, > best
              i - internal, S stale
Origin codes: i - IGP, e - EGP, ? - incomplete
Network             Next Hop          Metric             LocPrf Weight Path
*  10.0.0.0         172.16.232.55          0             0 300 88 90 99 ?
*>                  172.16.232.52       2222             0 400 ?
*  172.16.0.0       172.16.232.55          0             0 300 90 99 88 200 ?
*>                  172.16.232.52       2222             0 400 ?
* 192.168.199.0     172.16.232.55          0             0 300 88 90 99 ?
*>                  172.16.232.52       2222             0 400 ?

This table describes the significant fields shown in the display.

Table 14. show bgp inconsistent-as Field Descriptions

Field

Description

BGP router identifier

BGP identifier for the local system.

local AS number

Autonomous system number for the local system.

BGP main routing table version

Last version of the BGP database that was installed into the main routing table.

Dampening enabled

Displayed if dampening is enabled for the routes in this BGP routing table.

BGP scan interval

Interval (in seconds) between scans of the BGP table specified by the address family and subaddress family.

Status codes

Status of the table entry. The status is displayed as a three-character field at the beginning of each line in the table. The first character may be (in order of precedence):

S—Path is stale, indicating that a graceful restart is in progress with the peer from which the route was learned.

s—Path is more specific than a locally sourced aggregate route and has been suppressed.

*—Path is valid.

The second character may be (in order of precedence):

>—Path is the best path to use for that network.

d—Path is dampened.

h—Path is a history entry, representing a route that is currently withdrawn, but that is being maintained to preserve dampening information. Such routes should never be marked as valid.

The third character may be:

i—Path was learned by an internal BGP (iBGP) session.

Origin codes

Origin of the path. The origin code is displayed at the end of each line in the table. It can be one of the following values:

i—Path originated from an Interior Gateway Protocol (IGP) and was advertised with a network or aggregate-address command.

e—Path originated from an Exterior Gateway Protocol (EGP).

?—Origin of the path is not clear. Usually, this is a route that is redistributed into BGP from an IGP.

Network

IP prefix and prefix length for a network.

Next Hop

IP address of the next system that is used when a packet is forwarded to the destination network. An entry of 0.0.0.0 indicates that the router has a non-BGP route to this network.

Metric

Value of the interautonomous system metric, otherwise known as the Multi Exit Discriminator (MED) metric.

LocPrf

Local preference value. This is used to determine the preferred exit point from the local autonomous system. It is propagated throughout the local autonomous system.

Weight

Path weight. Weight is used in choosing the preferred path to a route. It is not advertised to any neighbor.

Path

Autonomous system path to the destination network. At the end of the path is the origin code for the path.

show bgp labels

To display Border Gateway Protocol (BGP) routes and their incoming and outgoing labels, use the show bgp labels command in EXEC mode.

show bgp labels

Syntax Description

ipv4

(Optional) Specifies IP Version 4 address prefixes.

unicast

(Optional) Specifies unicast address prefixes.

multicast

(Optional) Specifies multicast address prefixes.

labeled-unicast

(Optional) Specifies labeled-unicast address prefixes.

all

(Optional) For subaddress families, specifies prefixes for all subaddress families.

ipv6

(Optional) Specifies IP Version 6 address prefixes.

vpnv4 unicast

(Optional) Specifies VPNv4 unicast address families.

rd rd-address

(Optional) Displays routes with a specific route distinguisher.

vrf

(Optional) Specifies VPN routing and forwarding (VRF) instance.

vrf-name

(Optional) Name of a VRF.

all

(Optional) For VRF, specifies all VRFs.

ipv4 {unicast | labeled-unicast }

(Optional) For VRF, specifies IPv4 unicast or labeled-unicast address families.

ipv6 unicast

(Optional) For VRF, specifies IPv6 unicast address families.

Command Default

If no address family or subaddress family is specified, the default address family and subaddress family specified using the set default-afi and set default-safi commands are used.

Command Modes

EXEC

Command History

Release

Modification

Release 3.7.2

This command was introduced.

Usage Guidelines

To use this command, you must be in a user group associated with a task group that includes appropriate task IDs. If the user group assignment is preventing you from using a command, contact your AAA administrator for assistance.

Task ID

Task ID

Operations

bgp

read

Examples

The following is sample output from the show bgp labels command in EXEC mode:


RP/0/RSP0/CPU0:router# show bgp vrf BAR ipv4 unicast labels 

BGP VRF BAR, state: Active BGP Route Distinguisher: 100:1 BGP router identifier 10.1.1.1, local AS number 100 
BGP table state: Active BGP main routing table version 12

Status codes: s suppressed, d damped, h history, * valid, > best
              i - internal, S stale
Origin codes: i - IGP, e - EGP, ? - incomplete
   Network            Next Hop        Rcvd Label      Local Label
Route Distinguisher: 100:1 (default for vrf BAR)
*> 20.1.1.1/32        10.0.101.1      16              nolabel
*> 20.1.1.2/32        10.0.101.1      16              nolabel
*> 20.1.1.3/32        10.0.101.1      16              nolabel
*> 20.1.1.4/32        10.0.101.1      16              nolabel
*> 20.1.1.5/32        10.0.101.1      16              nolabel

Processed 5 prefixes, 5 paths

This table describes the significant fields shown in the display.

Table 15. show bgp labels Field Descriptions

Field

Description

BGP Route Distinguisher

BGP route distinguisher.

BGP router identifier

BGP identifier for the local system.

local AS number

Autonomous system number for the local system.

BGP table state

State of the BGP routing table.

BGP main routing table version

Last version of the BGP database that was installed into the main routing table.

Status codes

Status of the table entry. The status is displayed as a three-character field at the beginning of each line in the table. The first character may be (in order of precedence):

S—Path is stale, indicating that a graceful restart is in progress with the peer from which the route was learned.

s—Path is more specific than a locally sourced aggregate route and has been suppressed.

*—Path is valid.

The second character may be (in order of precedence):

>—Path is the best path to use for that network.

d—Path is dampened.

h—Path is a history entry, representing a route that is currently withdrawn, but that is being maintained to preserve dampening information. Such routes should never be marked as valid.

The third character may be:

i—Path was learned by an internal BGP (iBGP) session.

Origin codes

Origin of the path. The origin code is displayed at the end of each line in the table. It can be one of the following values:

i—Path originated from an Interior Gateway Protocol (IGP) and was sourced by BGP using a network or aggregate-address command.

e—Path originated from an Exterior Gateway Protocol (EGP).

?—Origin of the path is not clear. Usually, this is a route that is redistributed into BGP from an IGP.

Network

IP prefix and prefix length for a network.

Next Hop

IP address of the next system that is used when a packet is forwarded to the destination network. An entry of 0.0.0.0 indicates that the router has a non-BGP route to this network.

Rcvd Label

Received label.

Local Label

Local label.

show bgp l2vpn

To display BGP routes associated with VPLS or VPWS or EVPN under L2VPN address family, use theshow bgp l2vpn command in EXEC mode.

show bgp l2vpn { vpls | vpws | evpn } { recent-prefix [ private] | rd rd _value bgp_prefix[ detail ] }

Syntax Description

vpls

Specifies Virtual Private LAN Services (VPLS).

vpws

Specifies Virtual Private Wire Service (VPWS).

evpn

Specifies Ethernet Virtual Private Network (EVPN).

recent-prefixes

(Optional) Specifies summary of recently updated or deleted prefixes.

private

(Optional) Specifies private information.

rd _value

Value of the route distinguisher.

bgp_prefix

Specifies BGP prefix.

detail

Provides detailed output for the specified route distinguisher and BGP prefix.

Command Default

No default behavior or values

Command Modes

EXEC mode

EXEC

Command History

Release Modification
Release 3.9.1

This command was introduced.

Release 7.1.1

This command was modified. The keywords recent-prefixes and private were added.

Usage Guidelines

To use this command, you must be in a user group associated with a task group that includes appropriate task IDs. If the user group assignment is preventing you from using a command, contact your AAA administrator for assistance.

Task ID

Task ID Operation
bgp read

Examples

The following example is sample output from the show bgp l2vpn vpls for route distinguisher 1:1:

RP/0/RSP0/CPU0:router#show bgp l2vpn vpls rd 1:1 2:1
BGP routing table entry for 2:1/32, Route Distinguisher: 1:1
Versions:
  Process           bRIB/RIB  SendTblVer
  Speaker                  2           2
    Local Label: 16000
Paths: (1 available, best #1)
  Advertised to update-groups (with more than one peer):
    0.1 
  Advertised to peers (in unique update groups):
    100.100.100.1 
  Path #1: Received by speaker 0
  Local
    0.0.0.0 from 0.0.0.0 (200.200.200.1)
      Origin IGP, localpref 100, valid, redistributed, best, import-candidate
      Extended community: RT:4:4 L2VPN:19:0:1500 
      Block Size:10

The following example is sample output from the show bgp l2vpn vpws for route distinguisher 200:200:

RP/0/RSP0/CPU0:router#show bgp l2vpn vpws rd 200:200 3:1
BGP routing table entry for 3:1/32, Route Distinguisher: 200:200
Versions:
  Process           bRIB/RIB  SendTblVer
  Speaker                  6           6
    Local Label: 16015
Paths: (1 available, best #1)
  Advertised to update-groups (with more than one peer):
    0.1 
  Advertised to peers (in unique update groups):
    100.100.100.1 
  Path #1: Received by speaker 0
  Local
    0.0.0.0 from 0.0.0.0 (200.200.200.1)
      Origin IGP, localpref 100, valid, redistributed, best, import-candidate
      Extended community: RT:2:2 L2VPN:4:0:1500 
      Circuit Vector:0xfd 0xff 
      Block Size:10

The following example is sample output from the show bgp l2vpn vpls for local NLRI:

RD is 3.3.3.3:3276, NH Address is 100.0.0.1, and VPLS ID is 150:200. The RT is 200:100.


Note


The RT and VPLS-ID are always same for the same VPLS instance.


RP/0/RSP0/CPU0:router#show bgp l2vpn vpls
Sat Jun  6 17:01:18.610 PST
BGP router identifier 3.3.3.3, local AS number 101
BGP generic scan interval 60 secs
BGP table state: Active
Table ID: 0x0
BGP main routing table version 5
BGP scan interval 60 secs

Status codes: s suppressed, d damped, h history, * valid, > best
              i - internal, S stale
Origin codes: i - IGP, e - EGP, ? - incomplete
   Network            Next Hop 
Route Distinguisher: 3.3.3.3:3276 (default for vrf g1:b1)
*>i200.0.0.1          30.0.0.2        
*>i100.0.0.1          0.0.0.0         
Route Distinguisher: 2.2.2.2:3435
*>i200.0.0.1          30.0.0.2        

Processed 3 prefixes, 3 paths

RP/0/RSP0/CPU0:router#show bgp l2vpn vpls rd 3.3.3.3:3276  100.0.0.1
Sat Jun  6 16:40:03.191 PST
BGP routing table entry for 100.0.0.1, Route Distinguisher: 3.3.3.3:3276
Versions:
  Process           bRIB/RIB  SendTblVer
  Speaker                  3           3
    Last Modified: Jun  6 11:20:57.944 for 05:19:05
Paths: (1 available, best #1)
  Advertised to peers (in unique update groups):
    30.0.0.2 
  Path #1: Received by speaker 0
  Local
    0.0.0.0 from 0.0.0.0 (3.3.3.3)
      Origin IGP, localpref 100, valid, redistributed, best, import-candidate
      Extended community: RT:200:100 VPLS-ID:150:200

The following example is sample output from the show bgp l2vpn vpls for remote NLRI:

RD is 2.2.2.2:3435, NH Address is 200.0.0.1, and VPLS ID is 150:200. The RT is 200:100.
RP/0/RSP0/CPU0:router#show bgp l2vpn vpls rd 2.2.2.2:3435 200.0.0.1
Sat Jun  6 16:53:55.726 PST
BGP routing table entry for 200.0.0.1, Route Distinguisher: 2.2.2.2:3435
Versions:
  Process           bRIB/RIB  SendTblVer
  Speaker                  5           5
Last Modified: Jun  6 11:20:57.944 for 05:32:58
Paths: (1 available, best #1)
  Not advertised to any peer
  Path #1: Received by speaker 0
  Local
    30.0.0.2 from 30.0.0.2 (133.133.133.133)
      Origin IGP, localpref 100, valid, internal, best, import-candidate, imported
      Extended community: RT:200:50 VPLS-ID:150:200
The following example is sample output of the show bgp l2vpn evpn recent-prefixes command. This sample output provides details for the route distinguishers and corresponding prefix.
Router# show bgp l2vpn evpn  recent-prefixes 

Address-Family   Route-Distinguisher Prefix
================ =================== ==============
L2VPN EVPN       0:0:0               [5][0][32][198.51.100.22]/24
L2VPN EVPN       10.5.0.1:100        [5][0][32][192.0.2.1]/24
L2VPN EVPN       10.5.0.1:100        [5][0][32][192.0.2.2]/24
L2VPN EVPN       10.5.0.1:100        [5][0][32][192.0.2.3]/24
L2VPN EVPN       10.5.0.1:100        [5][0][32][192.0.2.4]/24
The following example is sample output of the show bgp l2vpn evpn rd command for a Data Center Interconnect Layer 3 Gateway. This sample output provides details for the specified route distinguisher and prefix.
RP/0/RSP0/CPU0:router# show bgp l2vpn evpn rd 100:1 [5][4231][112][fec0::1001]/176 detail
Fri Aug 21 00:34:43.747 PDT
BGP routing table entry for [5][4231][112][fec0::1001]/176, Route Distinguisher: 100:1
Versions:
  Process           bRIB/RIB  SendTblVer
  Speaker                  5           5
    Flags: 0x04040001+0x00000000; 
Last Modified: Aug 21 00:16:58.000 for 00:17:46
Paths: (1 available, best #1)
  Not advertised to any peer
  Path #1: Received by speaker 0
  Flags: 0x4000600025060005, import: 0x3f
  Not advertised to any peer
  Local
    11.0.0.1 (metric 2) from 20.0.0.1 (11.0.0.1)
      Received Label 16001
      Origin IGP, localpref 100, valid, internal, best, group-best, import-candidate, reoriginate, not-in-vrf
      Received Path ID 0, Local Path ID 1, version 5
      Extended community: Flags 0x2: Encapsulation Type:8 Router MAC:aabb.ccdd.eeff RT:65540:1 RT:40.40.40.40:1 RT:100:1 
      Originator: 11.0.0.1, Cluster list: 20.20.20.20
      EVPN ESI: ffff.ffff.ffff.ffff.ff01, Gateway Address : fec0::254

show bgp l2vpn vpls

To display L2VPN information on BGP summary, routes for a specified bridge group domain, advertised routes, routes with a specific route distinguisher, BGP neighbor connections, nexthops, and BGP process, use the show bgp l2vpn vpls command in EXEC mode.

show bgp l2vpn vpls {summary | rd | neighbors | nexthops | bdomain | advertised | process}

Syntax Description

summary

Displays the summary of BGP neighbor status.

rd

Displays routes with a specific route distinguisher.

neighbors

Displays detailed information on TCP and BGP neighbor connections.

nexthops

Shows nexthop related information.

bdomain

Displays routes for a specified Bridge Group:domain

advertised

Shows advertised routes.

process

Displays BGP process information.

Command Default

No default behavior or values

Command Modes

EXEC

Command History

Release Modification
Release 3.9.1

This command was introduced.

Usage Guidelines

To use this command, you must be in a user group associated with a task group that includes appropriate task IDs. If the user group assignment is preventing you from using a command, contact your AAA administrator for assistance.

Task ID

Task ID Operation
bgp

read

Examples

The following example is sample output from the show bgp l2vpn vpls command:


RP/0/RSP0/CPU0:router#show bgp l2vpn vpls 
Wed Mar 17 15:26:29.433 EDT
BGP router identifier 60.60.60.60, local AS number 1
BGP generic scan interval 60 secs
Non-stop routing is enabled
BGP table state: Active
Table ID: 0x0
BGP main routing table version 24001
BGP NSR Initial initsync version 1 (Reached)
BGP scan interval 60 secs

Status codes: s suppressed, d damped, h history, * valid, > best
              i - internal, S stale
Origin codes: i - IGP, e - EGP, ? - incomplete
   Network            Next Hop        Rcvd Label      Local Label
Route Distinguisher: 101:1 (default for vrf bg1:bg1_bd1)
*>i10.10.10.10/32     10.10.10.10     nolabel         nolabel         
*> 60.60.60.60/32     0.0.0.0         nolabel         nolabel         
Route Distinguisher: 102:1 (default for vrf bg1:bg1_bd2)
*>i10.10.10.10/32     10.10.10.10     nolabel         nolabel         
*> 60.60.60.60/32     0.0.0.0         nolabel         nolabel

The following example is sample output from the show bgp l2vpn vpls command with the summary keyword:


RP/0/RSP0/CPU0:router#show bgp l2vpn vpls summary 
Wed Mar 17 15:27:09.502 EDT
BGP router identifier 60.60.60.60, local AS number 1
BGP generic scan interval 60 secs
Non-stop routing is enabled
BGP table state: Active
Table ID: 0x0
BGP main routing table version 24001
BGP NSR Initial initsync version 1 (Reached)
BGP scan interval 60 secs

BGP is operating in STANDALONE mode.


Process       RcvTblVer   bRIB/RIB   LabelVer  ImportVer  SendTblVer  StandbyVer
Speaker           24001      24001      24001      24001       24001           0

Neighbor        Spk    AS MsgRcvd MsgSent   TblVer  InQ OutQ  Up/Down  St/PfxRcd
10.10.10.10       0     1   45532    8392    24001    0    0 03:06:25       8000

The following example is sample output from the show bgp l2vpn vpls command for Route Distinguisher: 101:1:


RP/0/RSP0/CPU0:router#show bgp l2vpn vpls rd 101:1
Wed Mar 17 15:27:31.347 EDT
BGP router identifier 60.60.60.60, local AS number 1
BGP generic scan interval 60 secs
Non-stop routing is enabled
BGP table state: Active
Table ID: 0x0
BGP main routing table version 24001
BGP NSR Initial initsync version 1 (Reached)
BGP scan interval 60 secs

Status codes: s suppressed, d damped, h history, * valid, > best
              i - internal, S stale
Origin codes: i - IGP, e - EGP, ? - incomplete
   Network            Next Hop        Rcvd Label      Local Label
Route Distinguisher: 101:1 (default for vrf bg1:bg1_bd1)
*>i10.10.10.10/32     10.10.10.10     nolabel         nolabel         
*> 60.60.60.60/32     0.0.0.0         nolabel         nolabel         

Processed 2 prefixes, 2 paths

The following example is sample output from the show bgp l2vpn vpls command for BGP neighbor 10.10.10.10:


RP/0/RSP0/CPU0:router#show bgp l2vpn vpls neighbors 10.10.10.10        
Wed Mar 17 15:28:28.766 EDT

BGP neighbor is 10.10.10.10
 Remote AS 1, local AS 1, internal link
 Remote router ID 10.10.10.10
  BGP state = Established, up for 03:07:44
  NSR State: None
  Last read 00:00:31, Last read before reset 00:00:00
  Hold time is 180, keepalive interval is 60 seconds
  Configured hold time: 180, keepalive: 60, min acceptable hold time: 3
  Last write 00:00:45, attempted 19, written 19
  Second last write 00:01:45, attempted 19, written 19
  Last write before reset 00:00:00, attempted 0, written 0
  Second last write before reset 00:00:00, attempted 0, written 0
  Last write pulse rcvd  Mar 17 15:27:57.362 last full not set pulse count 847
  Last write pulse rcvd before reset 00:00:00
  Socket not armed for io, armed for read, armed for write
  Last write thread event before reset 00:00:00, second last 00:00:00
  Last KA expiry before reset 00:00:00, second last 00:00:00
  Last KA error before reset 00:00:00, KA not sent 00:00:00
  Last KA start before reset 00:00:00, second last 00:00:00
  Precedence: internet
  Non-stop routing is enabled
  Graceful restart is enabled
  Restart time is 300 seconds
  Stale path timeout time is 1200 seconds
  Neighbor capabilities:
    Route refresh: advertised and received
    Graceful Restart (GR Awareness): received
    4-byte AS: advertised and received
    Address family IPv4 Unicast: advertised and received
    Address family VPNv4 Unicast: advertised and received
    Address family L2VPN VPLS: advertised and received
  Received 45533 messages, 0 notifications, 0 in queue
  Sent 8393 messages, 0 notifications, 0 in queue
  Minimum time between advertisement runs is 0 secs

 For Address Family: IPv4 Unicast
  BGP neighbor version 1
  Update group: 0.2
  AF-dependant capabilities:
    Graceful Restart Capability advertised and received
      Local restart time is 300, RIB purge time is 900 seconds
      Maximum stalepath time is 1200 seconds
      Remote Restart time is 300 seconds
  Route refresh request: received 0, sent 0
  0 accepted prefixes, 0 are bestpaths
  Cumulative no. of prefixes denied: 0. 
  Prefix advertised 0, suppressed 0, withdrawn 0
  Maximum prefixes allowed 524288
  Threshold for warning message 75%, restart interval 0 min
  An EoR was received during read-only mode
  Last ack version 1, Last synced ack version 0
  Outstanding version objects: current 0, max 0

 For Address Family: VPNv4 Unicast
  BGP neighbor version 1
  Update group: 0.2
  AF-dependant capabilities:
    Graceful Restart Capability advertised and received
      Local restart time is 300, RIB purge time is 900 seconds
      Maximum stalepath time is 1200 seconds
      Remote Restart time is 300 seconds
  Route refresh request: received 0, sent 0
  0 accepted prefixes, 0 are bestpaths
  Cumulative no. of prefixes denied: 0. 
  Prefix advertised 0, suppressed 0, withdrawn 0
  Maximum prefixes allowed 524288
  Threshold for warning message 75%, restart interval 0 min
  An EoR was received during read-only mode
  Last ack version 1, Last synced ack version 0
  Outstanding version objects: current 0, max 0

 For Address Family: L2VPN VPLS
  BGP neighbor version 24001
  Update group: 0.2
  AF-dependant capabilities:
    Graceful Restart Capability advertised and received
      Local restart time is 300, RIB purge time is 900 seconds
      Maximum stalepath time is 1200 seconds
      Remote Restart time is 300 seconds
  Route refresh request: received 0, sent 203
  8000 accepted prefixes, 8000 are bestpaths
  Cumulative no. of prefixes denied: 18172. 
    No policy: 0, Failed RT match: 18172
    By ORF policy: 0, By policy: 0
  Prefix advertised 8000, suppressed 0, withdrawn 0
  Maximum prefixes allowed 524288
  Threshold for warning message 75%, restart interval 0 min
  An EoR was received during read-only mode
  Last ack version 24001, Last synced ack version 0
  Outstanding version objects: current 0, max 2

  Connections established 1; dropped 0
  Local host: 60.60.60.60, Local port: 179
  Foreign host: 10.10.10.10, Foreign port: 50472
  Last reset 00:00:00

The following example is sample output from the show bgp l2vpn vpls command with the advertised keyword:


RP/0/RSP0/CPU0:router#show bgp l2vpn vpls advertised           
Wed Mar 17 15:29:13.787 EDT
Route Distinguisher: 101:1
60.60.60.60/32 is advertised to 10.10.10.10
  Path info:
    neighbor: Local           neighbor router id: 60.60.60.60
    valid  redistributed  best  import-candidate  
  Attributes after inbound policy was applied:
    next hop: 0.0.0.0
    EXTCOMM 
    origin: IGP  
    aspath: 
    extended community: RT:101:1 L2VPN AGI:1:101 

The following example is sample output from the show bgp l2vpn vpls command with the nexthops keyword:


RP/0/RSP0/CPU0:router#show bgp l2vpn vpls nexthops   
Wed Mar 17 15:29:36.357 EDT
Total Nexthop Processing
  Time Spent: 0.000 secs

Maximum Nexthop Processing
  Received: 82y46w
  Bestpaths Deleted: 0
  Bestpaths Changed: 0
  Time Spent: 0.000 secs

Last Notification Processing
  Received: 03:10:50
  Time Spent: 0.000 secs

Gateway Address Family: IPv4 Unicast
Table ID: 0xe0000000
Nexthop Count: 2
Critical Trigger Delay: 3000msec
Non-critical Trigger Delay: 10000msec

Nexthop Version: 1, RIB version: 1

Status codes: R/UR Reachable/Unreachable
              C/NC Connected/Not-connected
              L/NL Local/Non-local
              I    Invalid (Policy Match Failed)
Next Hop        Status          Metric        Notf      LastRIBEvent RefCount
10.10.10.10     [R][NC][NL]          2         1/0    03:10:50 (Cri)     8000/8003

The following example is sample output from the show bgp l2vpn vpls command with the process keyword:


RP/0/RSP0/CPU0:router#show bgp l2vpn vpls process  
Wed Mar 17 15:29:56.086 EDT

BGP Process Information: 
BGP is operating in STANDALONE mode
Autonomous System number format: ASPLAIN
Autonomous System: 1
Router ID: 60.60.60.60 (manually configured)
Default Cluster ID: 60.60.60.60
Active Cluster IDs:  60.60.60.60
Fast external fallover enabled
Neighbor logging is enabled
Enforce first AS enabled
Default local preference: 100
Default keepalive: 60
Graceful restart enabled
Restart time: 180
Stale path timeout time: 1200
RIB purge timeout time: 900
Non-stop routing is enabled
Update delay: 600
Generic scan interval: 60

Address family: L2VPN VPLS
Dampening is not enabled
Client reflection is enabled in global config
Scan interval: 60
Main Table Version: 24001
Table version synced to RIB: 1
RIB has not converged: version 0

Node                Process     Nbrs Estb Rst Upd-Rcvd Upd-Sent Nfn-Rcv Nfn-Snt
node0_RSP0_CPU0     Speaker        1    1   2    45347      237       0       0

show bgp memory history

To show the history of memory changes when memory reaches a threshold of more than 1% of rlimit (rlimit is the maximum memory usage of bgp before triggering bgp process self restart protection), use the show bgp memory history command in EXEC mode.


Note


This command applies to all the Cisco IOS XR products running BGP.

show bgp memory history

Syntax Description

This command has no keywords or arguments.

Command Default

None

Command Modes

EXEC

Command History

Release

Modification

Release 7.11.1

Release 7.3.6

This command was introduced.

Usage Guidelines

Use this command to monitor BGP memory history usage versus prefixes, paths, path elements (= best paths or advertised paths) and bgp attributes. If bgp memory usage keeps increasing over time while network items scale remains stable, this might be an indication of memory leak which needs further investigation.

Task ID

Task ID

Operations

bgp

read

Examples

The following example uses the show bgp memory history command to display syslogs when memory reaches the rlimit threshold.
**************************************************************************

RP/0/RP0/CPU0:Router#show bgp memory history

History of memory changes recorded for a threshold greater than 1.0% of rlimit.
Last shown record displays current values.
Network information for default VRF.

Time               Memory(MB)  Rlimit(%)  Memory diff(MB)  Networks  Paths     PathElems  Attributes
Oct  2 16:30:37    152         1          152              400       400       400        9
Oct  2 16:31:37    343         4          191              396952    396869    396952     725
Oct  2 16:32:37    425         5          81               524567    513979    524567     8408 
...
Oct  3 21:39:49    818         9          277              1297568   1347868   1297555    11109
Oct  3 21:40:49    1081        13         263              1974078   2414610   1974065    11111
Oct  4 00:19:10    1081        13         -1               1972907   2413813   1972894    11154 

show bgp neighbor-group

To display information about the Border Gateway Protocol (BGP) configuration for neighbor groups, use the show bgp neighbor-group command in EXECEXEC mode mode.

show bgp neighbor-group group-name {configuration [defaults] [nvgen] | inheritance | users}

Syntax Description

group-name

Name of the address family group to display.

configuration

(Optional) Displays the effective configuration for the neighbor group, including any configuration inherited by this neighbor group.

defaults

(Optional) Displays all configuration, including default configuration.

nvgen

(Optional) Displays output in show running-config command output.

If the defaults keyword is also specified, the output is not suitable for cutting and pasting into a configuration session.

inheritance

Displays the af-groups, session groups, and neighbor groups from which this neighbor group inherits configuration.

users

Displays the neighbors and neighbor groups that inherit configuration from this neighbor group.

Command Default

No default behavior or value

Command Modes

EXECEXEC mode

Command History

Release

Modification

Release 3.7.2

This command was introduced.

Usage Guidelines

To use this command, you must be in a user group associated with a task group that includes appropriate task IDs. If the user group assignment is preventing you from using a command, contact your AAA administrator for assistance.

Use the show bgp neighbor-group command with the group-name configuration argument and keyword to display the effective configuration of a neighbor group, including any configuration inherited from session groups, address family groups, and neighbor groups through application of the use command. The source of each configured command is also displayed.

Use the defaults keyword to display all configuration for the neighbor group, including default configuration. The command output identifies default onfiguration. Use the nvgen keyword to display configuration in the output form of show running-config command. Output in this form is suitable for cutting and pasting into a configuration session.

Theshow bgp neighbor-group command with the group-name inheritance argument and keyword displays the session groups, address family groups, and neighbor groups from which the specified neighbor group inherits configuration.

The show bgp neighbor-group group-name command displays the neighbors and neighbor groups that inherit configuration from the specified neighbor group.

Task ID

Task ID

Operations

bgp

read

Examples

The examples use the following configuration:


 af-group group3 address-family ipv4 unicast
  remove-private-AS
  soft-reconfiguration inbound
!
 af-group group2 address-family ipv4 unicast
  use af-group group3
  send-community-ebgp
  send-extended-community-ebgp
  capability orf prefix both
 !
 session-group group3
  dmzlink-bw
 !
 neighbor-group group3
  use session-group group3
  timers 30 90
 !
 neighbor-group group1
  remote-as 1982
  use neighbor-group group2
  address-family ipv4 unicast
  !
 !
 neighbor-group group2
  use neighbor-group group3
  address-family ipv4 unicast
   use af-group group2
   weight 100
  !

The following is sample output from the show bgp neighbor-group command with the configuration keyword:


Router# show bgp neighbor-group group1 configuration 

neighbor-group group1
 remote-as 1982                   []
 timers 30 90                     [n:group2 n:group3]
 dmzlink-bw                       [n:group2 n:group3 s:group3]
 address-family ipv4 unicast      []
  capability orf prefix both      [n:group2 a:group2]
  remove-private-AS               [n:group2 a:group2 a:group3]
  send-community-ebgp             [n:group2 a:group2]
  send-extended-community-ebgp    [n:group2 a:group2]
  soft-reconfiguration inbound    [n:group2 a:group2 a:group3]
  weight 100                      [n:group2]

The configuration source is shown to the right of each command. In the output, the remote-as command is configured directly on neighbor group group1, and the send-community-ebgp command is inherited from neighbor group group2, which in turn inherits the setting from af-group group2.

The following is sample output from the show bgp neighbor-group command with the users keyword. This output shows that the group1 neighbor group inherits session (address family-independent configuration parameters) from the group2 neighbor group. The group1 neighbor group also inherits IPv4 unicast configuration parameters from the group2 neighbor group:


Router# show bgp neighbor-group group2 users

Session:      n:group1
IPv4 Unicast: n:group1

The following is sample output from the show bgp neighbor-group command with the inheritance keyword. This output shows that the specified neighbor group group1 inherits session (address family-independent configuration) from neighbor group group2, which inherits its own session from neighbor group group3. Neighbor group group3 inherited its session from session group group3. It also shows that the group1 neighbor-group inherits IPv4 unicast configuration parameters from the group2 neighbor group, which in turn inherits them from the group2 af-group, which itself inherits them from the group3 af-group:


Router# show bgp neighbor-group group1 inheritance 

Session:      n:group2 n:group3 s:group3
IPv4 Unicast: n:group2 a:group2 a:group3

This table describes the significant fields shown in the display.

Table 16. show bgp neighbor-group Field Descriptions

Field

Description

[ ]

Configures the command directly on the specified address family group.

s:

Indicates the name that follows is a session group.

a:

Indicates the name that follows is an address family group.

n:

Indicates the name that follows is a neighbor group.

[dflt]

Indicates the setting is not explicitly configured or inherited, and the default value for the setting is used. This field may be shown when the defaults keyword is specified.

<not set>

Indicates that the default is for the setting to be disabled. This field may be shown when the defaults keyword is specified.

show bgp neighbors

To display information about Border Gateway Protocol (BGP) connections to neighbors, use the show bgp neighbors command in EXEC mode.

show bgp neighbors [performance-statistics | missing-eor] [standby]

show bgp neighbors ip-address [advertised-routes | dampened-routes | flap-statistics | performance-statistics | received | | {prefix-filter | routes} | routes] [standby]

show bgp neighbors ip-address [configuration | [defaults] | | nvgen | inheritance] [standby]

show bgp neighbors ip-address decoded-message-log [ [in | out] [standby]]

Syntax Description

ipv4

(Optional) Specifies IP Version 4 address prefixes.

unicast

(Optional) Specifies unicast address prefixes.

multicast

(Optional) Specifies multicast address prefixes.

labeled-unicast

(Optional) Specifies labeled unicast address prefixes.

all

(Optional) For subaddress families, specifies prefixes for all subaddress families.

tunnel

(Optional) Specifies tunnel address prefixes.

ipv6

(Optional) Specifies IP Version 6 address prefixes.

all

(Optional) For address family, specifies prefixes for all address families.

vpnv4 unicast

(Optional) Specifies VPNv4 unicast address families.

vrf

(Optional) Specifies VPN routing and forwarding (VRF) instance.

vrf-name

(Optional) Name of a VRF.

all

(Optional) For VRF, specifies all VRFs.

ipv4 { unicast | labeled-unicast }

(Optional) For VRF, specifies IPv4 unicast or labeled-unicast address families.

ipv6 unicast

(Optional) For VRF, specifies IPv6 unicast address families.

performance-statistics

(Optional) Displays performance statistics relative to work done by the BGP process for this neighbor.

missing-eor

(Optional) Displays neighbors that did not send end-of-rib (EoR) notification in read-only mode.

ip-address

(Optional) IP address of the BGP-speaking neighbor. If you omit this argument, all neighbors are displayed.

advertised-routes

(Optional) Displays all routes the router advertised to the neighbor.

dampened-routes

(Optional) Displays the dampened routes that are learned from the neighbor.

flap-statistics

(Optional) Displays flap statistics of the routes learned from the neighbor.

received { prefix-filter | routes }

(Optional) Displays information received from the BGP neighbor. The options are:

prefix-filter— Displays the prefix list filter.

routes—Displays routes from the neighbor before inbound policy

routes

(Optional) Displays routes learned from the neighbor.

configuration

(Optional) Displays the effective configuration for the neighbor, including any settings that have been inherited from session groups, neighbor groups, or af-groups used by this neighbor.

defaults

(Optional) Displays all configuration settings, including any default settings.

nvgen

(Optional) Displays output in the show running-config command output.

inheritance

(Optional) Displays the session groups, neighbor groups, and af-groups from which this neighbor inherits configuration settings.

decoded-message-log

(Optional) Displays BGP message logs.

in

(Optional) Displays BGP inbound messages.

out

(Optional) Displays BGP outbound messages.

standby

Displays standby BGP information.

detail

Displays detailed BGP information.

Command Default

If no address family or subaddress family is specified, the default address family and subaddress family specified using the set default-afi and set default-safi commands are used.

Command Modes

EXEC

Command History

Release

Modification

Release 7.9.1

This command displays the slow peer configuration state and slow peer detection or processing information.

Release 3.7.2

This command was introduced.

Release 3.9.0

Asplain format for 4-byte autonomous system numbers notation was supported. The input parameters and output were modified to display 4-byte autonomous system numbers and extended communities in either asplain or asdot notations.

Release 5.3.2

The command was modified to include graceful maintenance feature information.

Release 5.3.2

The command output was modified display TCP MSS information.

Release 5.3.2

The decoded-message-log [in | out] option was added.

Usage Guidelines


Note


The set default-afi command is used to specify the default address family for the session, and the set default-safi command is used to specify the default subaddress family for the session. See the System Management Command Reference for Cisco ASR 9000 Series Routers for detailed information and syntax for the set default-afi and set default-safi commands. If you do not specify a default address family, the default address family is IPv4. If you do not specify a default subaddress family, the default subaddress family is unicast.


BGP contains a separate routing table for each configured address family and subaddress family combination. The address family and subaddress family options specify which routing table should be examined. If the all keyword is specified for address family or subaddress family, each matching routing table is examined in turn.

Use the show bgp neighbors command to display detailed information about all neighbors or a specific neighbor. Use the performance-statistics keyword to display information about the work related to specific neighbors done by the BGP process.

Use the show bgp neighbors command with the ip-address received prefix-filter argument and keyword to display the Outbound Route Filter (ORF) received from a neighbor.

Use the advertised-routes keyword to display a summary of the routes advertised to the specified neighbor.

Use the dampened-routes keyword to display routes received from the specified neighbor that have been suppressed due to dampening. For more details, see the show bgp dampened-paths command.

To display information about flapping routes received from a neighbor, use the flap-statistics keyword. For more details, see the show bgp flap-statistics command.

To display the routes received from a neighbor, use the routes keyword. For more details, see the show bgp command.

Use the show bgp neighbor command with the ip-address configuration argument and keyword to display the effective configuration of a neighbor, including configuration inherited from session groups, neighbor groups, or af-groups through application of the use command. Use the defaults keyword to display the value of all configurations for the neighbor, including default configuration. Use the nvgen keyword to display configuration output format of the show running-config command. Output in this format is suitable for cutting and pasting into a configuration session. Use the show bgp neighbors command with the ip-address inheritance argument and keyword to display the session groups, neighbor groups, and af-groups from which the specified neighbor inherits configuration.

Task ID

Task ID

Operations

bgp

read

Examples

The following is the sample output from the show bgp neighbors command with the ip-address and configuration argument and keyword to display graceful maintenance feature attributes:
**************************************************************************

RP/0/0/CPU0:R1#show bgp neighbor 10.12.12.5
...
  Graceful Maintenance locally active, Local Pref=45, AS prepends=3
...
 For Address Family: IPv4 Unicast
...
  GSHUT Community attribute sent to this neighbor
...

**************************************************************************

RP/0/0/CPU0:R1#show bgp neighbor 10.12.12.5 configuration 
Mon Feb  2 14:30:41.042 PST
neighbor 12.12.12.5
 remote-as 1                         []
 graceful-maintenance 1              []
 gr-maint local-preference 45        []
 gr-maint as-prepends 3              []
 gr-maint activate                   []

**************************************************************************

Examples

The following is sample output from the show bgp neighbors command:


RP/0/RSP0/CPU0:router# show bgp neighbors 10.0.101.1

BGP neighbor is 10.0.101.1, remote AS 2, local AS 1, external link
 Description: routem neighbor
 Remote router ID 10.0.101.1
  BGP state = Established, up for 00:00:56
  TCP open mode: passive only
BGP neighbor is 1.1.1.2
 Remote AS 300, local AS 100, external link
 Remote router ID 0.0.0.0
  BGP state = Idle (LC/FIB for the neighbor in reloading)
  Last read 00:00:00, Last read before reset 00:05:12
  Hold time is 180, keepalive interval is 60 seconds
  Configured hold time: 180, keepalive: 60, min acceptable hold time: 3

  BFD enabled (session initializing)
  Last read 00:00:55, hold time is 180, keepalive interval is 60 seconds
  DMZ-link bandwidth is 1000 Mb/s
  Neighbor capabilities:
    Route refresh: advertised
    4-byte AS: advertised and received
    Address family IPv4 Unicast: advertised and received
    Address family IPv4 Multicast: advertised and received
  Received 119 messages, 0 notifications, 0 in queue
  Sent 119 messages, 22 notifications, 0 in queue
  Minimum time between advertisement runs is 60 seconds

 For Address Family: IPv4 Unicast
  BGP neighbor version 137
  Update group: 1.3
  Community attribute sent to this neighbor
  AF-dependant capabilities:
    Outbound Route Filter (ORF) type (128) Prefix-list:
      Send-mode: advertised
      Receive-mode: advertised
  Route refresh request: received 0, sent 0
  Policy for incoming advertisements is pass-all
  Policy for outgoing advertisements is pass-all
  5 accepted prefixes, 5 are bestpaths
  Prefix advertised 3, suppressed 0, withdrawn 0, maximum limit 1000000
  Threshold for warning message 75%


 For Address Family: IPv4 Multicast
  BGP neighbor version 23
  Update group: 1.2
  Route refresh request: received 0, sent 0
  Policy for incoming advertisements is pass-all
  Policy for outgoing advertisements is pass-all
  2 accepted prefixes, 2 are bestpaths
  Prefix advertised 0, suppressed 0, withdrawn 0, maximum limit 131072
  Threshold for warning message 75%

  Connections established 9; dropped 8
  Last reset 00:02:10, due to User clear requested (CEASE notification sent - administrative reset)
  Time since last notification sent to neighbor: 00:02:10
  Error Code: administrative reset
  Notification data sent:
    None

This table describes the significant fields shown in the display.

Table 17. show bgp neighbors Field Descriptions

Field

Description

BGP neighbor

IP address of the BGP neighbor and its autonomous system number. If the neighbor is in the same autonomous system as the router, then the link between them is internal; otherwise, it is considered external.

Description

Neighbor specific description.

remote AS

  • Number of the autonomous system to which the neighbor belongs.
  • Range for 2-byte Autonomous system numbers (ASNs) is 1 to 65535.
  • Range for 4-byte Autonomous system numbers (ASNs) in asplain format is 1 to 4294967295.
  • Range for 4-byte Autonomous system numbers (ASNs) is asdot format is 1.0 to 65535.65535.

local AS

Autonomous system number of the local system.

  • Range for 2-byte Autonomous system numbers (ASNs) is 1 to 65535.
  • Range for 4-byte Autonomous system numbers (ASNs) in asplain format is 1 to 4294967295.
  • Range for 4-byte Autonomous system numbers (ASNs) is asdot format is 1.0 to 65535.65535.

internal link

Neighbor is an internal BGP peer.

external link

Neighbor is an external BGP peer.

Administratively shut down

Neighbor connection is disabled using the shutdown command.

remote router ID

Router ID (an IP address) of the neighbor.

Neighbor under common administration

Neighbor is internal or a confederation peer.

BGP state

Internal state of this BGP connection.

BFD enabled

Status of bidirectional forwarding detection.

TCP open mode

TCP mode used in establishing the BGP session. The following valid TCP mode are supported:

  • default—Accept active/passive connections
  • passive-only—Accept only passive connections
  • active-only—Accept only active connections initiated by the router

Last read

Time since BGP last read a message from this neighbor.

hold time

Hold time (in seconds) used on the connection with this neighbor.

keepalive interval

Interval for sending keepalives to this neighbor.

DMZ-link bandwidth

DMZ link bandwidth for this neighbor.

Neighbor capabilities

BGP capabilities advertised and received from this neighbor. The following valid BGP capabilities are supported:

  • Multi-protocol
  • Route refresh
  • Graceful restart
  • Outbound Route Filter (ORF) type (128) Prefix

Route refresh

Indicates that the neighbor supports dynamic soft reset using the route refresh capability.

4-byte AS

Indicates that the neighbor supports the 4-byte AS capability.

Address family

Indicates that the local system supports the displayed address family capability. If “received” is displayed, the neighbor also supports the displayed address family.

Received

Number of messages received from this neighbor, the number of notification messages received and processed from this neighbor, and the number of messages that have been received, but not yet processed.

Sent

Number of messages sent to this neighbor, the number of notification messages generated to be sent to this neighbor, and the number of messages queued to be sent to this neighbor.

Minimum time between advertisement runs

Advertisement interval (in seconds) for this neighbor.

For Address Family

Information that follows is specific to the displayed address family.

BGP neighbor version

Last version of the BGP database that was sent to the neighbor for the specified address family.

Update group

Update group to which the neighbor belongs.

Route reflector client

Indicates that the local system is acting as a route reflector for this neighbor.

Inbound soft reconfiguration allowed

Indicates that soft reconfiguration is enabled for routes received from this neighbor.

Note

 

If the neighbor has route refresh capability, then soft configuration received-only routes are not stored by the local system unless “override route refresh” is displayed.

eBGP neighbor with no inbound or outbound policy: defaults to drop

Indicates that the neighbor does not have an inbound or outbound policy configured using the route-policy (BGP) command. Hence, no routes are accepted from or advertised to this neighbor.

Private AS number removed from updates to this neighbor

Indicates that remove-private-AS is configured on the specified address family for this neighbor.

NEXT_HOP is always this router

Indicates that next-hop-self is configured on the specified address family for this neighbor.

Community attribute sent to this neighbor

Indicates that send-community-ebgp is configured on the specified address family for this neighbor.

Extended community attribute sent to this neighbor

Indicates that send-extended-community-ebgp is configured on the specified address family for this neighbor.

Default information originate

Indicates that default-originate is configured on the specified address family for this neighbor, together with the policy used, if one was specified in the default-originate configuration. An indication of whether the default route has been advertised to the neighbor is also shown.

AF-dependant capabilities

BGP capabilities that are specific to a particular address family. The following valid AF-dependent BGP capabilities are supported:

  • route refresh capability
  • route refresh capability OLD value

Outbound Route Filter

Neighbor has the Outbound Route Filter (ORF) capability for the specified address family. Details of the capabilities supported are also shown:

Send-mode—“advertised” is shown if the local system can send an outbound route filter to the neighbor. “received” is shown if the neighbor can send an outbound route filter to the local system.

Receive-mode—“advertised” is shown if the local system can receive an outbound route filter from the neighbor. “received” is shown if the neighbor can receive an outbound route filter from the local system.

Graceful Restart Capability

Indicates whether graceful restart capability has been advertised to and received from the neighbor for the specified address family.

Neighbor preserved the forwarding state during latest restart

Indicates that when the neighbor connection was last established, the neighbor indicated that it preserved its forwarding state for the specified address family.

Local restart time

Restart time (in seconds) advertised to this neighbor.

RIB purge time

RIB purge time (in seconds) used for graceful restarts.

Maximum stalepath time

Maximum time (in seconds) a path received from this neighbor may be marked as stale if the neighbor restarts.

Remote Restart time

Restart time received from this neighbor.

Route refresh request

Number of route refresh requests sent and received from this neighbor.

Outbound Route Filter (ORF)

“sent” indicates that an outbound route filter has been sent to this neighbor. “received” indicates that an outbound route filter has been received from this neighbor.

Note

 

A received outbound route filter may be displayed using the show bgp neighbors command with the received prefix-filter keywords.

First update is deferred until ORF or ROUTE-REFRESH is received

If the local system advertised the receive capability and the neighbor has advertised send capability, no updates are generated until specifically asked by the neighbor (using a ROUTE-REFRESH or ORF with immediate request).

Scheduled to send the Prefix-list filter

Indicates the local system is due to send an outbound route filter request in order to receive updates from the neighbor.

Inbound path policy

Indicates if an inbound path policy is configured.

Outbound path policy

Indicates if an outbound path policy is configured.

Incoming update prefix filter list

Indicates a prefix list is configured to filter inbound updates from the neighbor.

Default weight

Default weight for routes received from the neighbor.

Policy for incoming advertisements

Indicates a route policy is configured to be applied to inbound updates from the neighbor.

Policy for outgoing advertisements

Indicates a route policy is configured to be applied to outbound updates to the neighbor.

Type

Indicates whether the condition map selects routes that should be advertised, or routes that should not be advertised:

Exist—Routes advertised if permitted by the condition route map.

Non-exist—Routes advertised if denied by the condition route map.

accepted prefixes

Number of prefixes accepted.

Prefix advertised

Number of prefixes advertised to the neighbor during the lifetime of the current connection with the neighbor.

suppressed

Number of prefix updates that were suppressed because no transitive attributes changed from one best path to the next.

Note

 

Update suppression occurs only for external BGP neighbors.

withdrawn

Number of prefixes withdrawn from the neighbor during the lifetime of the current connection with the neighbor.

maximum limit

Maximum number of prefixes that may be received from the neighbor. If “(warning-only)” is displayed, a warning message is generated when the limit is exceeded, otherwise the neighbor connection is shut down when the limit is exceeded.

Threshold for warning message

Percentage of maximum prefix limit for the neighbor at which a warning message is generated.

Connections established

Number of times the router has established a BGP peering session with the neighbor.

dropped

Number of times that a good connection has failed or been taken down.

Last reset due to

Reason that the connection with the neighbor was last reset.

Time since last notification sent to neighbor

Amount of time since a notification message was last sent to the neighbor.

Error Code

Type of notification that was sent. The notification data, if any, is also displayed.

Time since last notification received from neighbor

Amount of time since a notification message was last received from the neighbor.

Error Code

Type of notification that was received. The notification data received, if any, is also displayed

External BGP neighbor may be up to <n> hops away

Indicates ebgp-multihop is configured for the neighbor.

External BGP neighbor not directly connected

Indicates that the neighbor is not directly attached to the local system.

Notification data sent:

Data providing more details on the error along with the error notification sent to the neighbor.

The following is sample output from the show bgp neighbors command with the advertised-routes keyword:


RP/0/RSP0/CPU0:router# show bgp neighbors 172.20.16.178 routes

BGP router identifier 172.20.16.181, local AS number 1
BGP main routing table version 27
BGP scan interval 60 secs
Status codes: s suppressed, d damped, h history, * valid, > best
              i - internal, S stale
Origin codes: i - IGP, e - EGP, ? - incomplete
Network             Next Hop             Metric LocPrf Weight Path
*> 10.0.0.0         172.20.16.178        40             0 10 ?
*> 10.22.0.0        172.20.16.178        40             0 10 ?

The following is sample output from the show bgp neighbors command with the routes keyword:


RP/0/RSP0/CPU0:router# show bgp neighbors 10.0.101.1 dampened-routes 

BGP router identifier 10.0.0.5, local AS number 1
BGP main routing table version 48
Dampening enabled
BGP scan interval 60 secs
Status codes: s suppressed, d damped, h history, * valid, > best
              i - internal, S stale
Origin codes: i - IGP, e - EGP, ? - incomplete
   Network            From            Reuse    Path
*d 10.0.0.0           10.0.101.1      00:59:30 2 100 1000 i
*d 11.0.0.0           10.0.101.1      00:59:30 2 100 1000 i
*d 12.0.0.0           10.0.101.1      00:59:30 2 100 1000 i
*d 13.0.0.0           10.0.101.1      00:59:30 2 100 1000 i
*d 14.0.0.0           10.0.101.1      00:59:30 2 100 1000 i

This table describes the significant fields shown in the display.

Table 18. show bgp neighbors routes Field Descriptions

Field

Description

BGP router identifier

BGP identifier for the local system.

local AS number

Autonomous system number for the local system.

BGP main routing table version

Last version of the BGP database that was installed into the main routing table.

Dampening enabled

Displayed if dampening is enabled for the routes in this BGP routing table.

BGP scan interval

Interval (in seconds) between scans of the BGP table specified by the address family and subaddress family.

Status codes

Status of the table entry. The status is displayed as a three-character field at the beginning of each line in the table. The first character may be (in order of precedence):

S—Path is stale, indicating that a graceful restart is in progress with the peer from which the route was learned.

s—Path is more specific than a locally sourced aggregate route and has been suppressed.

*—Path is valid.

The second character may be (in order of precedence):

>—Path is the best path to use for that network.

d—Path is dampened.

h—Path is a history entry, representing a route that is currently withdrawn, but that is being maintained to preserve dampening information. Such routes should never be marked as valid.

The third character may be:

i—Path was learned by an internal BGP (iBGP) session.

Origin codes

Origin of the path. The origin code is displayed at the end of each line in the table. It can be one of the following values:

i—Path originated from an Interior Gateway Protocol (IGP) and was advertised with a network or aggregate-address command.

e—Path originated from an Exterior Gateway Protocol (EGP).

?—Origin of the path is not clear. Usually, this is a route that is redistributed into BGP from an IGP.

Network

IP prefix and prefix length for a network.

Next Hop

IP address of the next system that is used when a packet is forwarded to the destination network. An entry of 0.0.0.0 indicates that the router has a non-BGP route to this network.

Metric

Value of the interautonomous system metric, otherwise known as the Multi Exit Discriminator (MED) metric.

LocPrf

Local preference value. This is used to determine the preferred exit point from the local autonomous system. It is propagated throughout the local autonomous system.

Weight

Path weight. Weight is used in choosing the preferred path to a route. It is not advertised to any neighbor.

Path

Autonomous system path to the destination network. At the end of the path is the origin code for the path.

The following is sample output from the show bgp neighbors command with the dampened-routes keyword:


RP/0/RSP0/CPU0:router# show bgp neighbors 10.0.101.1 flap-statistics

BGP router identifier 10.0.0.5, local AS number 1
BGP main routing table version 48
Dampening enabled
BGP scan interval 60 secs
Status codes: s suppressed, d damped, h history, * valid, > best
              i - internal, S stale
Origin codes: i - IGP, e - EGP, ? - incomplete
   Network            From            Flaps Duration Reuse    Path
 h 10.1.0.0           10.0.101.1      5008  2d02h             2 5000 1000 
 h 10.2.0.0           10.0.101.1      5008  2d02h             2 2000 3000 
 h 10.2.0.0           10.0.101.1      5008  2d02h             2 9000 6000 
*d 10.0.0.0           10.0.101.1      5008  2d02h    00:59:30 2 100 1000 
 h 10.0.0.0/16        10.0.101.1      5008  2d02h             2 100 102 
*d 10.11.0.0          10.0.101.1      5008  2d02h    00:59:30 2 100 1000 
*d 10.12.0.0          10.0.101.1      5008  2d02h    00:59:30 2 100 1000 
*d 10.13.0.0          10.0.101.1      5008  2d02h    00:59:30 2 100 1000 
*d 10.14.0.0          10.0.101.1      5008  2d02h    00:59:30 2 100 1000 
 h 192.168.0.0/16     10.0.101.1      5008  2d02h             2 100 101 

This table describes the significant fields shown in the display.

Table 19. show bgp neighbors dampened-routes Field Descriptions

Field

Description

BGP router identifier

BGP identifier for the local system.

local AS number

Autonomous system number for the local system.

BGP main routing table version

Last version of the BGP database that was installed into the main routing table.

Dampening enabled

Displayed if dampening is enabled for the routes in this BGP routing table.

BGP scan interval

Interval (in seconds) between scans of the BGP table specified by the address family and subaddress family.

Status codes

Status of the table entry. The status is displayed as a three-character field at the beginning of each line in the table. The first character may be (in order of precedence):

S—Path is stale, indicating that a graceful restart is in progress with the peer from which the route was learned.

s—Path is more specific than a locally sourced aggregate route and has been suppressed.

*—Path is valid.

The second character may be (in order of precedence):

>—Path is the best path to use for that network.

d—Path is dampened.

h—Path is a history entry, representing a route that is currently withdrawn, but that is being maintained to preserve dampening information. Such routes should never be marked as valid.

The third character may be:

i—Path was learned by an internal BGP (iBGP) session.

Origin codes

Origin of the path. The origin code is displayed at the end of each line in the table. It can be one of the following values:

i—Path originated from an Interior Gateway Protocol (IGP) and was advertised with a network or aggregate-address command.

e—Path originated from an Exterior Gateway Protocol (EGP).

?—Origin of the path is not clear. Usually, this is a route that is redistributed into BGP from an IGP.

Network

IP prefix and prefix length for a network.

From

Neighbor from which the route was received.

Reuse

Time (in hours:minutes:seconds) after which the path is made available.

Path

Autonomous system path to the destination network. At the end of the path is the origin code for the path.

The following is sample output from the show bgp neighbors command with the flap-statistics keyword:


RP/0/RSP0/CPU0:router# show bgp neighbors 10.0.101.2 performance-statistics 

BGP neighbor is 10.0.101.2, remote AS 1
  Read 3023 messages (58639 bytes) in 3019 calls (time spent: 1.312 secs)
  Read throttled 0 times
  Processed 3023 inbound messages (time spent: 0.198 secs)
  Wrote 58410 bytes in 6062 calls (time spent: 3.041 secs)
  Processing write list: wrote 0 messages in 0 calls (time spent: 0.000 secs)
  Processing write queue: wrote 3040 messages in 3040 calls (time spent: 0.055 secs)

  Received 3023 messages, 0 notifications, 0 in queue
  Sent 3040 messages, 0 notifications, 0 in queue

This table describes the significant fields shown in the display.

Table 20. show bgp neighbors flap-statistics Field Descriptions

Field

Description

BGP route identifier

BGP identifier for the local system.

local AS number

Autonomous system number for the local system.

BGP main routing table version

Last version of the BGP database that was installed into the main routing table.

Dampening enabled

Displayed if dampening has been enabled for the routes in this BGP routing table.

BGP scan interval

Interval (in seconds) between when the BGP process scans for the specified address family and subaddress family.

Status codes

Status of the table entry. The status is displayed as a three-character field at the beginning of each line in the table. The first character may be (in order of precedence):

S—Path is stale, indicating that a graceful restart is in progress with the peer from which the route was learned.

s—Path is more specific than a locally sourced aggregate route and has been suppressed.

*—Path is valid.

The second character may be (in order of precedence):

d—Path is dampened.

h—Path is a history entry, representing a route that is currently withdrawn, but that is being maintained to preserve dampening information. Such routes should never be marked as valid.

The third character may be:

i—Path was learned by an internal BGP (iBGP) session.

Origin codes

Origin of the path. The origin code is displayed at the end of each line in the table. It can be one of the following values:

i—Path originated from an Interior Gateway Protocol (IGP) and was advertised with a network command.

e—Path originated from an Exterior Gateway Protocol (EGP).

?—Origin of the path is not clear. Usually, this is a route that is redistributed into BGP from an IGP.

Network

IP prefix and prefix length for a network.

From

IP address of the peer that advertised this route.

Flaps

Number of times the route has flapped.

Duration

Time (in hours:minutes:seconds) since the router noticed the first flap.

Reuse

Time (in hours:minutes:seconds) after which the path is made available.

Path

Autonomous system path to reach the destination network.

The following is sample output from the show bgp neighbors command with the performance-statistics keyword:


RP/0/RSP0/CPU0:router# show bgp neighbors 10.0.101.2 performance-statistics 
BGP neighbor is 10.0.101.2, remote AS 1
  Read 3023 messages (58639 bytes) in 3019 calls (time spent: 1.312 secs)
  Read throttled 0 times
  Processed 3023 inbound messages (time spent: 0.198 secs)
  Wrote 58410 bytes in 6062 calls (time spent: 3.041 secs)
  Processing write list: wrote 0 messages in 0 calls (time spent: 0.000 secs)
  Processing write queue: wrote 3040 messages in 3040 calls (time spent: 0.055 secs)
  Received 3023 messages, 0 notifications, 0 in queue
  Sent 3040 messages, 0 notifications, 0 in queue

This table describes the significant fields shown in the display.

Table 21. show bgp neighbors performance-statistics Field Descriptions

Field

Description

Read

Indicates the number of messages received from the neighbor, the total size of received messages, the number of read operations performed, and the real time spent (in seconds) by the process performing read operations for this neighbor.

Read throttled

Number of times that reading from the TCP connection to this neighbor has been throttled. Throttling is due to a backlog of messages that have been read but not processed.

inbound messages

Number of read messages that have been processed, and the real time spent processing inbound messages for this neighbor.

Wrote

Amount of data that has been sent to this neighbor, number of write operations performed, and the real time spent by the process performing write operations for this neighbor.

Processing write list

Number of messages written from the write list to this neighbor, number of times the write list has been processed, and real time spent processing the write list.

Note

 

Write lists typically contain only update messages.

Processing write queue

Number of messages written from the write queue to this neighbor, number of times the write queue has been processed, and real time spent processing the write queue.

Received

Number of messages received from this neighbor, number of notification messages received and processed from this neighbor, and number of messages that have been received, but not yet processed.

Sent

Number of messages sent to this neighbor, number of notification messages generated to be sent to this neighbor, and number of messages queued to be sent to this neighbor.

The following is sample output from the show bgp neighbors command with the configuration keyword:


RP/0/RSP0/CPU0:router# show bgp neighbors 10.0.101.1 configuration

neighbor 10.0.101.1
 remote-as 2                   []
 bfd fast-detect               []
 address-family ipv4 unicast   []
  policy pass-all in           []
  policy pass-all out          []
 address-family ipv4 multicast []
  policy pass-all in           []
  policy pass-all out          []

This table describes the significant fields shown in the display.

Table 22. show bgp neighbors configuration Field Descriptions

Field

Description

neighbor

IP address configuration of the neighbor.

remote-as

Remote autonomous system configured on the neighbor.

bfd fast-detect

BFD parameter configured on the neighbor.

address-family

Address family and subsequent address family configured on the router.

route-policy pass-all in

Route policy configured for inbound updates.

route-policy pass-all out

Route policy configured for outbound updates.

This sample output from the show bgp neighbor command displays TCP MSS information for the specified neighbor:



RP/0/RSP0/CPU0:router#show bgp neighbor 10.0.0.2

BGP neighbor is 10.0.0.2
Remote AS 1, local AS 1, internal link
Remote router ID 10.0.0.2
BGP state = Established, up for 00:09:17
Last read 00:00:16, Last read before reset 00:00:00
Hold time is 180, keepalive interval is 60 seconds
Configured hold time: 180, keepalive: 60, min acceptable hold time: 3
Last write 00:00:16, attempted 19, written 19
Second last write 00:01:16, attempted 19, written 19
Last write before reset 00:00:00, attempted 0, written 0
Second last write before reset 00:00:00, attempted 0, written 0
Last write pulse rcvd Dec 7 11:58:42.411 last full not set pulse count 23
Last write pulse rcvd before reset 00:00:00
Socket not armed for io, armed for read, armed for write
Last write thread event before reset 00:00:00, second last 00:00:00
Last KA expiry before reset 00:00:00, second last 00:00:00
Last KA error before reset 00:00:00, KA not sent 00:00:00
Last KA start before reset 00:00:00, second last 00:00:00
Precedence: internet
Multi-protocol capability received
Neighbor capabilities:
Route refresh: advertised (old + new) and received (old + new)
Graceful Restart (GR Awareness): advertised and received
4-byte AS: advertised and received
Address family IPv4 Unicast: advertised and received
Received 12 messages, 0 notifications, 0 in queue
Sent 12 messages, 0 notifications, 0 in queue
Minimum time between advertisement runs is 0 secs
TCP Maximum Segment Size 500 

For Address Family: IPv4 Unicast
BGP neighbor version 4
Update group: 0.2 Filter-group: 0.1 No Refresh request being processed
Route refresh request: received 0, sent 0
0 accepted prefixes, 0 are bestpaths
Cumulative no. of prefixes denied: 0. 
Prefix advertised 0, suppressed 0, withdrawn 0
Maximum prefixes allowed 1048576
Threshold for warning message 75%, restart interval 0 min
AIGP is enabled
An EoR was received during read-only mode
Last ack version 4, Last synced ack version 0
Outstanding version objects: current 0, max 0
Additional-paths operation: None
Send Multicast Attributes

This sample output from the show bgp neighbor command with the configuration keyword displays TCP MSS configuration:



RP/0/RSP0/CPU0:router#show bgp neighbor 10.0.0.2 configuration 

neighbor 10.0.0.2
remote-as 1 []
tcp-mss 400 [n:n1]
address-family IPv4 Unicast []

The sample output from the show bgp neighbor with the detail keyword displays the neighbor address-family slow peer configuration state and slow peer detection or processing information.
Router# show bgp neighbors 198.51.100.254 detail
Thu Dec  1 02:40:40.301 UTC

BGP neighbor is 198.51.100.254
Remote AS 1, local AS 1, internal link
Remote router ID 198.51.100.254
Cluster ID 209.165.201.1
  BGP state = Established, up for 04:31:32
  Previous State: Active
  Last Received Message: Update
  NSR State: None
...
For Address Family: IPv6 Labeled-unicast
  BGP neighbor version 0
  Update group: 0.2 Filter-group: 0.60  No Refresh request being processed
  Route-Reflector Client
  Inbound soft reconfiguration allowed (override route-refresh)
  AF-dependent capabilities:
    Additional-paths Send: advertised
    Additional-paths Receive: advertised
  Route refresh request: received 0, sent 0
  Slow Peer State: Dynamic
    Detected state: TRUE, Detection threshold: 120
    Detection Count: 3, Recovery Count: 2
    Processing slow peer: FALSE
    Dynamic Trigger Count: 163, Dynamic Process Count: 1

Where:

  • Slow Peer State indicates the effective configuration state (considering both global and neighbor address-family configuration) of neighbor address-family. Slow Peer State can be Static or Dynamic or Detection-only or None.

  • Detected State indicates if the slow peer event is triggered for this neighbor address-family. TRUE if the neighbor address-family slow peer event is triggered, else FALSE.

  • Detection threshold is the effective threshold configured (considering both global and neighbor address-family configuration) for the neighbor address-family. Threshold is used in determining if the peer (neighbor address-family) is slow.

  • Processing slow peer indicates slow peer processing state of the neighbor address-family. TRUE if neighbor address-family is processed as slow peer else FALSE.

show bgp neighbors nsr

To display Border Gateway Protocol (BGP) nonstop routing (NSR) information across neighbors, use the show bgp neighbors nsr command in EXEC mode.

show bgp [ipv4 {unicast | multicast | all} | ipv6 {unicast | multicast | all} | vpnv4 unicast | vpnv6 unicast | vrf {all | vrf_name}] neighbors nsr [standby]

Syntax Description

ipv4

(Optional) Specifies IP Version 4 address prefixes.

ipv6

(Optional) Specifies IP Version 6 address prefixes.

unicast

(Optional) Specifies unicast address prefixes.

multicast

(Optional) Specifies multicast address prefixes.

all

(Optional) For address family, specifies prefixes for all address families.

vpnv4 unicast

(Optional) Specifies VPNv4 unicast address families.

vpnv6 unicast

(Optional) Specifies VPNv6 unicast address families.

vrf

(Optional) Specifies VPN routing and forwarding (VRF) instance.

vrf_name

(Optional) Name of a VRF.

all

(Optional) For VRF, specifies all VRFs.

standby

(Optional) Displays information about the standby card.

Command Default

No default behavior or values.

Command Modes

EXEC

Command History

Release

Modification

Release 3.9.0

This command was introduced.

Usage Guidelines

To use this command, you must be in a user group associated with a task group that includes appropriate task IDs. If the user group assignment is preventing you from using a command, contact your AAA administrator for assistance.

Task ID

Task ID

Operations

bgp

read

Examples

The following is sample output from the show bgp neighbors nsr command with the standby keyword:


RP/0/RSP0/CPU0:router# show bgp neighbors nsr standby

BGP neighbor is 2.2.2.2
  BGP state = Established, up for 5d04h
  NSR state = NSR Ready
  Outstanding Postits: 0
 

BGP neighbor is 10.0.101.5
  BGP state = Established, up for 05:19:00
  NSR state = NSR Ready
  Outstanding Postits: 0
 

BGP neighbor is 10.1.0.5
  BGP state = Established, up for 5d04h
  NSR state = NSR Ready
  Outstanding Postits: 0

This table describes the significant fields shown in the display.

Table 23. show bgp neighbors nsr Field Descriptions

Field

Description

BGP state

Displays BGP neighbor peering state.

NSR state

Displays BGP neighbor NSR state.

Outstanding Postits

Displays the postit counters of pending events.

show bgp nexthops

To display statistical information about the Border Gateway Protocol (BGP) next hops, use the show bgp nexthops command in EXEC mode.

show bgp nexthops A.B.C.D. aigp-value [statistics] [speaker speaker-id] []

Syntax Description

ipv4

(Optional) Specifies IP Version 4 address prefixes.

unicast

(Optional) Specifies unicast address prefixes.

multicast

(Optional) Specifies multicast address prefixes.

labeled-unicast

(Optional) Specifies labeled-unicast address prefixes.

all

(Optional) For subaddress families, specifies prefixes for all subaddress families.

tunnel

(Optional) Specifies tunnel address prefixes.

ipv6

(Optional) Specifies IP Version 6 address prefixes.

all

(Optional) For address family, specifies prefixes for all address families.

vpnv4 unicast

(Optional) Specifies VPNv4 unicast address families.

vrf

(Optional) Specifies VPN routing and forwarding (VRF) instance.

vrf-name

(Optional) Name of a VRF.

all

(Optional) For VRF, specifies all VRFs.

ipv4 { unicast | labeled-unicast }

(Optional) For VRF, specifies IPv4 unicast or labeled-unicast address families.

ipv6 unicast

(Optional) For VRF, specifies IPv6 unicast address families.

statistics

(Optional) Specifies nexthop statistics.

speaker speaker-id

(Optional) Specifies a speaker process ID.

A.B.C.D

Next hop to display information about

aigp-value

Displays next hop statistics

Command Default

No default behavior or value

Command Modes

EXEC

Command History

Release

Modification

Release 3.7.2

This command was introduced.

Usage Guidelines

To use this command, you must be in a user group associated with a task group that includes appropriate task IDs. If the user group assignment is preventing you from using a command, contact your AAA administrator for assistance.

The show bgp nexthops command displays statistical information about next-hop notifications, the time spent processing the notifications, and details about each next-hop that has been registered with the Routing Information Base (RIB).

Use the vrf vrf-name keyword and argument to display only the next-hops present in the specified VPN routing and forwarding (VRF) instance.

The next-hop information is displayed for all active speaker processes in distributed mode. Each speaker displays a set of next-hops that belongs to the prefixes received by the speaker and next hops that belong to best paths that were received by other speaker processes. Use the speaker speaker-id keyword and argument to display information for only the specified speaker process. The distributed mode must be defined using the distributed speaker command for the speaker keyword to be available.

Task ID

Task ID

Operations

bgp

read

Examples

The following is sample output from the show bgp nexthops command with the VRF specified:


RP/0/RSP0/CPU0:router# show bgp vrf all nexthops 

Fri Mar 13 17:05:40.656 UTC 
 
VRF: 900
========
 
Total Nexthop Processing
  Time Spent: 0.000 secs
 
Maximum Nexthop Processing
  Received: 82y48w
  Bestpaths Deleted: 0
  Bestpaths Changed: 0
  Time Spent: 0.000 secs
 
Last Notification Processing
  Received: 1d22h
  Time Spent: 0.000 secs
 
IPv4 Unicast is active
 
Gateway Address Family: IPv4 Unicast
Table ID: 0xe0000001
Nexthop Count: 2
Critical Trigger Delay: 0msec
Non-critical Trigger Delay: 10000msec
 
Nexthop Version: 1, RIB version: 1
 
Status codes: R/UR Reachable/Unreachable
              C/NC Connected/Not-connected
              L/NL Local/Non-local
              I    Invalid (Policy Match Failed)
Next Hop        Status          Metric        Notf      LastRIBEvent RefCount
10.0.101.201    [UR]        4294967295         0/0       1d22h (Reg)        0/3
90.0.0.2        [R][C][NL]           0         1/0       1d22h (Cri)       20/23
 

VRF: 901
========
 
Total Nexthop Processing
  Time Spent: 0.000 secs
 
Maximum Nexthop Processing
  Received: 82y48w
  Bestpaths Deleted: 0
  Bestpaths Changed: 0
  Time Spent: 0.000 secs
 
Last Notification Processing
  Received: 1d22h
  Time Spent: 0.000 secs
 
IPv4 Unicast is active
 
Gateway Address Family: IPv4 Unicast
Table ID: 0xe0000002
Nexthop Count: 2
Critical Trigger Delay: 0msec
Non-critical Trigger Delay: 10000msec
 
Nexthop Version: 1, RIB version: 1
 
Status codes: R/UR Reachable/Unreachable
              C/NC Connected/Not-connected
              L/NL Local/Non-local
              I    Invalid (Policy Match Failed)
Next Hop        Status          Metric        Notf      LastRIBEvent RefCount
10.0.101.201    [UR]        4294967295         0/0       1d22h (Reg)        0/3
91.0.0.2        [R][C][NL]           0         1/0       1d22h (Cri)       10/13
 

VRF: 902
========
 
Total Nexthop Processing
  Time Spent: 0.000 secs
 
Maximum Nexthop Processing
  Received: 82y48w
  Bestpaths Deleted: 0
  Bestpaths Changed: 0
  Time Spent: 0.000 secs
 
Last Notification Processing
  Received: 1d22h
  Time Spent: 0.000 secs
 
IPv4 Unicast is active
 
Gateway Address Family: IPv4 Unicast
Table ID: 0xe0000003
Nexthop Count: 2
Critical Trigger Delay: 0msec
Non-critical Trigger Delay: 10000msec
 
Nexthop Version: 1, RIB version: 1
 
Status codes: R/UR Reachable/Unreachable
              C/NC Connected/Not-connected
              L/NL Local/Non-local
              I    Invalid (Policy Match Failed)
Next Hop        Status          Metric        Notf      LastRIBEvent RefCount
10.0.101.201    [UR]        4294967295         0/0       1d22h (Reg)        0/3
92.0.0.2        [R][C][NL]           0         1/0       1d22h (Cri)       10/13
 

VRF: 903
========
 
Total Nexthop Processing
  Time Spent: 0.000 secs
 
Maximum Nexthop Processing
  Received: 82y48w
  Bestpaths Deleted: 0
  Bestpaths Changed: 0
  Time Spent: 0.000 secs
 
Last Notification Processing
  Received: 1d22h
  Time Spent: 0.000 secs
 
IPv4 Unicast is active
 
Gateway Address Family: IPv4 Unicast
Table ID: 0xe0000004
Nexthop Count: 2
Critical Trigger Delay: 0msec
Non-critical Trigger Delay: 10000msec
 
Nexthop Version: 1, RIB version: 1
 
Status codes: R/UR Reachable/Unreachable
              C/NC Connected/Not-connected
              L/NL Local/Non-local
              I    Invalid (Policy Match Failed)
Next Hop        Status          Metric        Notf      LastRIBEvent RefCount
10.0.101.201    [UR]        4294967295         0/0       1d22h (Reg)        0/3
93.0.0.2        [R][C][NL]           0         1/0       1d22h (Cri)       10/13
 

VRF: 904
========
 
Total Nexthop Processing
  Time Spent: 0.000 secs
 
Maximum Nexthop Processing
  Received: 82y48w
  Bestpaths Deleted: 0
  Bestpaths Changed: 0
  Time Spent: 0.000 secs
 
Last Notification Processing
  Received: 1d22h
  Time Spent: 0.000 secs
 
IPv4 Unicast is active
 
Gateway Address Family: IPv4 Unicast
Table ID: 0xe0000005
Nexthop Count: 2
Critical Trigger Delay: 0msec
Non-critical Trigger Delay: 10000msec
 
Nexthop Version: 1, RIB version: 1
 
Status codes: R/UR Reachable/Unreachable
              C/NC Connected/Not-connected
              L/NL Local/Non-local
              I    Invalid (Policy Match Failed)
Next Hop        Status          Metric        Notf      LastRIBEvent RefCount
10.0.101.201    [UR]        4294967295         0/0       1d22h (Reg)        0/3
94.0.0.2        [R][C][NL]           0         1/0       1d22h (Cri)       10/13
 

VRF: 905
========
 
Total Nexthop Processing
  Time Spent: 0.000 secs
 
Maximum Nexthop Processing
  Received: 82y48w
  Bestpaths Deleted: 0
  Bestpaths Changed: 0
  Time Spent: 0.000 secs
 
Last Notification Processing
  Received: 1d22h
  Time Spent: 0.000 secs
 
IPv4 Unicast is active
 
Gateway Address Family: IPv4 Unicast
Table ID: 0xe0000006
Nexthop Count: 2
Critical Trigger Delay: 0msec
Non-critical Trigger Delay: 10000msec
 
Nexthop Version: 1, RIB version: 1
 
Status codes: R/UR Reachable/Unreachable
              C/NC Connected/Not-connected
              L/NL Local/Non-local
              I    Invalid (Policy Match Failed)
Next Hop        Status          Metric        Notf      LastRIBEvent RefCount
10.0.101.201    [UR]        4294967295         0/0       1d22h (Reg)        0/3
95.0.0.2        [R][C][NL]           0         1/0       1d22h (Cri)       10/13
 

VRF: 906
========
 
Total Nexthop Processing
  Time Spent: 0.000 secs
 
Maximum Nexthop Processing
  Received: 82y48w
  Bestpaths Deleted: 0
  Bestpaths Changed: 0
  Time Spent: 0.000 secs
 
Last Notification Processing
  Received: 1d22h
  Time Spent: 0.000 secs
 
IPv4 Unicast is active
 
Gateway Address Family: IPv4 Unicast
Table ID: 0xe0000007
Nexthop Count: 2
Critical Trigger Delay: 0msec
Non-critical Trigger Delay: 10000msec
 
Nexthop Version: 1, RIB version: 1
 
Status codes: R/UR Reachable/Unreachable
              C/NC Connected/Not-connected
              L/NL Local/Non-local
              I    Invalid (Policy Match Failed)
Next Hop        Status          Metric        Notf      LastRIBEvent RefCount
10.0.101.201    [UR]        4294967295         0/0       1d22h (Reg)        0/3
96.0.0.2        [R][C][NL]           0         1/0       1d22h (Cri)       10/13
 

VRF: 907
========
 
Total Nexthop Processing
  Time Spent: 0.000 secs
 
Maximum Nexthop Processing
  Received: 82y48w
  Bestpaths Deleted: 0
  Bestpaths Changed: 0
  Time Spent: 0.000 secs
 
Last Notification Processing
  Received: 1d22h
  Time Spent: 0.000 secs
 
IPv4 Unicast is active
 
Gateway Address Family: IPv4 Unicast
Table ID: 0xe0000008
Nexthop Count: 2
Critical Trigger Delay: 0msec
Non-critical Trigger Delay: 10000msec
 
Nexthop Version: 1, RIB version: 1
 
Status codes: R/UR Reachable/Unreachable
              C/NC Connected/Not-connected
              L/NL Local/Non-local
              I    Invalid (Policy Match Failed)
Next Hop        Status          Metric        Notf      LastRIBEvent RefCount
10.0.101.201    [UR]        4294967295         0/0       1d22h (Reg)        0/3
97.0.0.2        [R][C][NL]           0         1/0       1d22h (Cri)       10/13
 

VRF: 908
========
 
Total Nexthop Processing
  Time Spent: 0.000 secs
 
Maximum Nexthop Processing
  Received: 82y48w
  Bestpaths Deleted: 0
  Bestpaths Changed: 0
  Time Spent: 0.000 secs
 
Last Notification Processing
  Received: 1d22h
  Time Spent: 0.000 secs
 
IPv4 Unicast is active
 
Gateway Address Family: IPv4 Unicast
Table ID: 0xe0000009
Nexthop Count: 2
Critical Trigger Delay: 0msec
Non-critical Trigger Delay: 10000msec
 
Nexthop Version: 1, RIB version: 1
 
Status codes: R/UR Reachable/Unreachable
              C/NC Connected/Not-connected
              L/NL Local/Non-local
              I    Invalid (Policy Match Failed)
Next Hop        Status          Metric        Notf      LastRIBEvent RefCount
10.0.101.201    [UR]        4294967295         0/0       1d22h (Reg)        0/3
98.0.0.2        [R][C][NL]           0         1/0       1d22h (Cri)       10/13
 

VRF: 909
========
 
Total Nexthop Processing
  Time Spent: 0.000 secs
 
Maximum Nexthop Processing
  Received: 82y48w
  Bestpaths Deleted: 0
  Bestpaths Changed: 0
  Time Spent: 0.000 secs
 
Last Notification Processing
  Received: 1d22h
  Time Spent: 0.000 secs
 
IPv4 Unicast is active
 
Gateway Address Family: IPv4 Unicast
Table ID: 0xe000000a
Nexthop Count: 1
Critical Trigger Delay: 0msec
Non-critical Trigger Delay: 10000msec
 
Nexthop Version: 1, RIB version: 1
 
Status codes: R/UR Reachable/Unreachable
              C/NC Connected/Not-connected
              L/NL Local/Non-local
              I    Invalid (Policy Match Failed)
Next Hop        Status          Metric        Notf      LastRIBEvent RefCount
99.0.0.2        [UR]        4294967295         0/0       1d22h (Reg)        0/3
 

VRF: yellow
===========
 
Total Nexthop Processing
  Time Spent: 0.000 secs
 
Maximum Nexthop Processing
  Received: 82y48w
  Bestpaths Deleted: 0
  Bestpaths Changed: 0
  Time Spent: 0.000 secs
 
Last Notification Processing
  Received: 82y48w
  Time Spent: 0.000 secs
 
IPv4 Unicast is active
 
Gateway Address Family: IPv4 Unicast
Table ID: 0xe000000e
Nexthop Count: 0
Critical Trigger Delay: 0msec
Non-critical Trigger Delay: 10000msec
 
Nexthop Version: 1, RIB version: 1

This table describes the significant fields shown in the display.

Table 24. show bgp vrf all nexthops Field Descriptions

Field

Description

VRF

Name of the VRF.

Total Nexthop Processing Time Spent

Time spent processing trigger delays for critical and noncritical events for the VRF or address family. The time is specified in seconds.

Maximum Nexthop Processing

Time that has passed since the nexthop notification was received that resulted in spending the maximum amount of processing time for all notifications.

Last Notification Processing

Time that has passed since the last nexthop notification was received.

IPv4 Unicast is active.

VRF specified output that indicates the IPv4 unicast address family is active within the VRF.

Nexthop Count

Number of next hops for the VRF or address family.

Critical Trigger Delay

Configured critical trigger delay.

Non-critical Trigger Delay

Configured noncritical trigger delay.

Total Critical Notifications Received

Number of critical notifications received.

Total Non-critical Notifications Received

Number of noncritical notifications received.

Bestpaths Deleted After Last Walk

Number of best paths deleted due to the last notification.

Bestpaths Changed After Last Walk

Number of best paths modified due to the last notification.

Next Hop

IP address of the next hop.

Status

Status of the next hop.

Metric

IGP metric of the next hop.

Notf

Number of critical and noncritical notifications received.

LastRIBEvent

When the last notification was received from the RIB.

RefCount

The number of neighbors or prefixes that refer to the next hop in address family/all format.

Address Family

Name of the address family.

show bgp nsr

To display Border Gateway Protocol (BGP) nonstop routing (NSR) information, use the show bgp nsr command in EXEC mode.

show bgp [ipv4 {unicast | multicast | labeled-unicast | all | tunnel} | ipv6 {unicast | multicast | all | labeled-unicast} | all {unicast | multicast | all | labeled-unicast | tunnel} | vpnv4 unicast | vrf {vrf-name | all} [ipv4 {unicast | labeled-unicast} | ipv6 unicast] | vpvn6 unicast] nsr [standby]

Syntax Description

ipv4

(Optional) Specifies IP Version 4 address prefixes.

unicast

(Optional) Specifies unicast address prefixes.

multicast

(Optional) Specifies multicast address prefixes.

labeled-unicast

(Optional) Specifies labeled unicast address prefixes.

all

(Optional) For subaddress families, specifies prefixes for all subaddress families.

tunnel

(Optional) Specifies tunnel address prefixes.

mdt

(Optional) Specifies multicast distribution tree (MDT) address prefixes.

multicast

(Optional) Specifies multicast address prefixes.

ipv6

(Optional) Specifies IP Version 6 address prefixes.

all

(Optional) For address family, specifies prefixes for all address families.

vpnv4 unicast

(Optional) Specifies VPNv4 unicast address families.

vrf

(Optional) Specifies VPN routing and forwarding (VRF) instance.

vrf-name

(Optional) Name of a VRF.

all

(Optional) For VRF, specifies all VRFs.

ipv4 { unicast | labeled-unicast }

(Optional) For VRF, specifies IPv4 unicast or labeled-unicast address families.

ipv6 unicast

(Optional) For VRF, specifies IPv6 unicast address families.

vpnv6 unicast

(Optional) Specifies VPNv6 unicast address families.

standby

Displays information about the standby card.

Command Default

If no address family or subaddress family is specified, the default address family and subaddress family specified using the set default-afi and set default-safi commands are used.

Command Modes

EXEC

Command History

Release

Modification

Release 3.9.0

This command was introduced.

Usage Guidelines

To use this command, you must be in a user group associated with a task group that includes appropriate task IDs. If the user group assignment is preventing you from using a command, contact your AAA administrator for assistance.

Task ID

Task ID

Operations

bgp

read

Examples

The following is sample output from the show bgp nsr command:


RP/0/RSP0/CPU0:router# show bgp nsr

Fri Jan 30 10:18:48.171 PST PDT

BGP Process Information: 
BGP is operating in STANDALONE mode
Autonomous System: 100
Router ID: 10.1.0.1 (manually configured)
Default Cluster ID: 10.1.0.1
Active Cluster IDs:  10.1.0.1
Fast external fallover enabled
Neighbor logging is not enabled
Enforce first AS enabled
AS Path ignore is enabled
AS Path multipath-relax is enabled
Default local preference: 100
Default keepalive: 60
Graceful restart enabled
Restart time: 180
Stale path timeout time: 360
RIB purge timeout time: 600
Non-stop routing is enabled
Update delay: 120
Generic scan interval: 60

Address family: IPv4 Unicast
Dampening is not enabled
Client reflection is enabled in global config
Scan interval: 60
Main Table Version: 7034
IGP notification: IGPs notified
RIB has converged: version 1

========== Post Failover Summary for Active instance ==========

Node                Process            Read      Write    Inbound


node0_0_CPU0        Speaker          146.75      18.90       3.46

  Entered mode  Standby Ready               : Jan 30 10:00:39
  Entered mode  TCP NSR Setup               : Jan 30 10:00:39
  Entered mode  TCP NSR Setup Done          : Jan 30 10:00:39
  Entered mode  TCP Initial Sync            : Jan 30 10:00:39
  Entered mode  TCP Initial Sync Done       : Jan 30 10:00:44
  Entered mode  FPBSN processing done       : Jan 30 10:00:44
  Entered mode  Update processing done      : Jan 30 10:00:44
  Entered mode  BGP Initial Sync            : Jan 30 10:00:44
  Entered mode  BGP Initial Sync done       : Jan 30 10:00:44
  Entered mode  NSR Ready                   : Jan 30 10:00:44

Current BGP NSR state - NSR Ready achieved at: Jan 30 10:00:44
NSR State READY notified to Redcon at: Jan 30 10:16:58

NSR Post Failover Summary:


QAD Statistics:

  Messages Sent       : 512            ACKs Received       : 512       
  Messages Received   : 8              ACKs Sent           : 8         
  Send Failures       : 1              Send ACK Failures   : 0         
  Suspends            : 1              Resumes             : 1         
  Messages Processed  : 8              Out of sequence drops: 0              

Postit Summary:

  Total pending postit messages:  0
  Neighbors with pending postits: 0



Conv  Bestpath   TunnelUpd  Import     RIBUpd     Label      ReadWrite  LastUpd  
Process: Speaker

Yes   120        ---        ---        120        120        120        87531    

Rib Trigger: enabled
Last RIB down event  Jan 29 09:50:03.069 received
Last RIB convergence Jan 29 09:50:03.069 last ack received.

Address Family IPv4 Unicast converged in 87531 seconds

The following example shows sample output from the show bgp nsr command with the standby keyword:


RP/0/RSP0/CPU0:router# show bgp nsr standby

Fri Jan 30 10:18:55.654 PST PDT

BGP Process Information: 
BGP is operating in STANDALONE mode
Autonomous System: 100
Router ID: 10.1.0.1 (manually configured)
Default Cluster ID: 10.1.0.1
Active Cluster IDs:  10.1.0.1
Fast external fallover enabled
Neighbor logging is not enabled
Enforce first AS enabled
AS Path ignore is enabled
AS Path multipath-relax is enabled
Default local preference: 100
Default keepalive: 60
Graceful restart enabled
Restart time: 180
Stale path timeout time: 360
RIB purge timeout time: 600
Non-stop routing is enabled
Update delay: 120
Generic scan interval: 60

Address family: IPv4 Unicast
Dampening is not enabled
Client reflection is enabled in global config
Scan interval: 60
Main Table Version: 7034
IGP notification: IGPs notified
RIB has converged: version 1

========== Post Failover Summary for Standby instance ==========

Node                Process            Read      Write    Inbound


node0_1_CPU0        Speaker            1.68       0.00       1.42

Entered mode  Standby Ready               : Jan 30 10:00:39
Entered mode  TCP Replication             : Jan 30 10:00:39
Entered mode  TCP Init Sync Done          : Jan 30 10:00:44
Entered mode  NSR Ready                   : Jan 30 10:00:44

QAD Statistics:

  Messages Sent       : 9              ACKs Received       : 9         
  Messages Received   : 512            ACKs Sent           : 512       
  Send Failures       : 0              Send ACK Failures   : 0         
  Suspends            : 0              Resumes             : 0         
  Messages Processed  : 512            Standby init drops  : 0         Out of sequence drops: 0              

Postit Summary:

  Total pending postit messages:  0
  Neighbors with pending postits: 0



Conv  Bestpath   TunnelUpd  Import     RIBUpd     Label      ReadWrite  LastUpd  
Process: Speaker

Yes   1233338444 ---        ---        1233338444 1233338444 1233338444 ---      

Rib Trigger: enabled
Last RIB down event  Jan 29 09:50:17.308 received
Last RIB convergence Jan 29 09:50:17.308 last ack received.

show bgp paths

To display all the Border Gateway Protocol (BGP) paths in the database, use the show bgp paths command in EXEC mode.

show bgp paths [detail] [debug] [regexp regular-expression]

Syntax Description

detail

(Optional) Displays detailed attribute information.

debug

(Optional) Displays attribute process ID, hash bucket, and hash chain ID attribute information.

regexp regular-expression

(Optional) Specifies an autonomous system path that matches the regular expression.

Command Default

No default behavior or values

Command Modes

EXEC

Command History

Release

Modification

Release 3.7.2

This command was introduced.

Release 3.9.0

Asplain format for 4-byte Autonomous system numbers notation was supported. The input parameters and output were modified to display 4-byte autonomous system numbers and extended communities in either asplain or asdot notations.

Usage Guidelines

To use this command, you must be in a user group associated with a task group that includes appropriate task IDs. If the user group assignment is preventing you from using a command, contact your AAA administrator for assistance.

Use the show bgp paths command to display information about AS paths and the associated attributes with which the paths were received.

If no options are specified, all stored AS paths are displayed with the number of routes using each path.


Note


The AS path information is stored independently of the address family, making it possible that routes from different address families could be using the same path.


Use the regular-expression argument to limit the output to only those paths that match the specified regular expression. See the Cisco ASR 9000 Series Aggregation Services Router Getting Started Guide for information on regular expressions.

Use the detail keyword to display detailed information on the attributes stored with the AS path.

Task ID

Task ID

Operations

bgp

read

Examples

The following is sample output from the show bgp paths command:


RP/0/RSP0/CPU0:router# show bgp paths detail 

Proc  Attributes                    Refcount    Metric Path
Spk 0 ORG AS LOCAL                         7          0 i
Spk 0 ORG AS LOCAL COMM EXTCOMM            3          0 21 i
Spk 0 MET ORG AS                           3         55 2 i
Spk 0 ORG AS                               3          0 2 10 11 i
Spk 0 ORG AS COMM                          3          0 2 10 11 i
Spk 0 MET ORG AS ATOM                      3          2 2 3 4 ?
Spk 0 MET ORG AS                           3          1 2 3 4 e
Spk 0 MET ORG AS                           3          0 2 3 4 i

This table describes the significant fields shown in the display.

Table 25. show bgp paths Field Descriptions

Field

Description

Proc

ID of the process in which the path is stored. This is always “Spk 0.”

Attributes

Attributes that are present. The following may appear:

MET —Multi Exit Discriminator (MED) attribute is present.

ORG—Origin attribute is present.

AS—AS path attribute is present.

LOCAL—Local preference attribute is present.

AGG—Aggregator attribute is present.

COMM—Communities attribute is present.

ATOM—Atomic aggregate attribute is present.

EXTCOMM—Extended communities attribute is present.

NeighborAS

Autonomous system number of the neighbor, or 0, if the path information originated locally.

  • Range for 2-byte Autonomous system numbers (ASNs) is 1 to 65535.
  • Range for 4-byte Autonomous system numbers (ASNs) in asplain format is 1 to 4294967295.
  • Range for 4-byte Autonomous system numbers (ASNs) is asdot format is 1.0 to 65535.65535.

Refcount

Number of routes using a path.

Metric

Value of the interautonomous system metric, otherwise known as the MED metric.

Path

Autonomous system path to the destination network. At the end of the path is the origin code for the path:

i—Path originated from an Interior Gateway Protocol (IGP) and was advertised with a network or aggregate-address command.

e—Path originated from an Exterior Gateway Protocol (EGP).

?—Origin of the path is not clear. Usually, this is a route that is redistributed into BGP from an IGP.

show bgp policy

To display information about Border Gateway Protocol (BGP) advertisements under a proposed policy, use the show bgp policy command in EXECEXEC mode mode.

show bgp policy

Syntax Description

ipv4

(Optional) Specifies IP Version 4 address prefixes.

unicast

(Optional) Specifies unicast address prefixes.

multicast

(Optional) Specifies multicast address prefixes.

labeled-unicast

(Optional) Specifies labeled unicast address prefixes.

all

(Optional) For subaddress families, specifies prefixes for all subaddress families.

tunnel

(Optional) Specifies tunnel address prefixes.

ipv6

(Optional) Specifies IP Version 6 address prefixes.

all

(Optional) For address family, specifies prefixes for all address families.

vpnv4 unicast

(Optional) Specifies VPNv4 unicast address families.

rd rd-address

(Optional) Displays routes with a specific route distinguisher.

vrf

(Optional) Specifies VPN routing and forwarding (VRF) instance.

vrf-name

(Optional) Name of a VRF.

all

(Optional) For VRF, specifies all VRFs.

ipv4 { unicast | labeled-unicast }

(Optional) For VRF, specifies IPv4 unicast or labeled-unicast address families.

ipv6 unicast

(Optional) For VRF, specifies IPv6 unicast address families.

neighbor

(Optional) Previews advertisements for a single neighbor.

ip-address

(Optional) IP address of a single neighbor.

sent-advertisements

(Optional) Displays the routes that have been advertised to neighbors. If a route has not yet been advertised to the neighbor, it is not shown.

route-policy

(Optional) Displays advertisements for an output route policy.

route-policy-name

(Optional) Name of the route policy.

summary

(Optional) Displays a summary of the BGP advertisements.

Command Default

Advertisements for all neighbors are displayed if the neighbor ip-address keyword and argument are not specified. If no address family or subaddress family is specified, the default address family and subaddress family specified using the set default-afi and set default-safi commands are used.

Command Modes

EXECEXEC mode

Command History

Release

Modification

Release 3.7.2

This command was introduced.

Usage Guidelines

To use this command, you must be in a user group associated with a task group that includes appropriate task IDs. If the user group assignment is preventing you from using a command, contact your AAA administrator for assistance.


Note


The set default-afi command is used to specify the default address family for the session, and the set default-safi command is used to specify the default subaddress family for the session. See the System Management Command Reference for Cisco ASR 9000 Series Routersfor detailed information and syntax for the set default-afi and set default-safi commands. If you do not specify a default address family, the default address family is IPv4. If you do not specify a default subaddress family, the default subaddress family is unicast.


BGP contains a separate routing table for each configured address family and subaddress family combination. The address family and subaddress family options specify the routing table to be examined. If the all keyword is specified for the address family or subaddress family, each matching routing table is examined in turn.

Use the show bgp policy command to display routes that would be advertised to neighbors under a proposed policy. Unlike in the show bgp advertised command, the information displayed reflects any modifications made to the routes when executing the specified policy.

Use the neighbor keyword to limit the output to routes advertised to a particular neighbor. Use the sent-advertisements keyword to change the output in two ways:

  • If a policy is not specified explicitly, any policy configured on the neighbor (using the route-policy (BGP) command) is executed before displaying the routes.
  • Only routes that have already been advertised to the neighbor (and not withdrawn) are displayed. Routes that have not yet been advertised are not displayed.

Use the summary keyword to display abbreviated output.

Task ID

Task ID

Operations

bgp

read

Examples

The following is sample output from the show bgp policy command with the summary keyword in EXECEXEC mode mode:


Router# show bgp policy summary

Network           Next Hop        From            Advertised to
172.16.1.0/24     10.0.101.1      10.0.101.1      10.0.101.2
                                                  10.0.101.3

172.17.0.0/16     0.0.0.0         Local           10.0.101.1
                                                  10.0.101.2
                                                  10.0.101.3

This table describes the significant fields shown in the display.

Table 26. show bgp policy summary Field Descriptions

Field

Description

Network

IP prefix and prefix length for a network.

Next Hop

IP address of the next system that is used when a packet is forwarded to the destination network. An entry of 0.0.0.0 indicates that the router has a non-BGP route to this network.

From

IP address of the peer that advertised this route.

Local

Indicates the route originated on the local system.

Local Aggregate

Indicates the route is an aggregate created on the local system.

Advertised to

Indicates the neighbors to which this route was advertised.

The following is sample output from the show bgp policy command in EXECEXEC mode mode:


Router# show bgp policy 

11.0.0.0/24 is advertised to 10.4.101.1
  Path info:
    neighbor: Local           neighbor router id: 10.4.0.1
    valid local best
  Attributes after inbound policy was applied:
    next hop: 0.0.0.0
    MET ORG AS
    origin: IGP metric: 0
    aspath:
  Attributes after outbound policy was applied:
    next hop: 10.4.0.1
    MET ORG AS
    origin: IGP metric: 0
    aspath: 1

11.0.0.0/24 is advertised to 10.4.101.2
  Path info:
    neighbor: Local           neighbor router id: 10.4.0.1
    valid local best
  Attributes after inbound policy was applied:
    next hop: 0.0.0.0
    MET ORG AS
    origin: IGP metric: 0
    aspath:
  Attributes after outbound policy was applied:
    next hop: 10.4.0.1
    MET ORG AS
    origin: IGP metric: 0
    aspath: 

11.0.0.0/24 is advertised to 10.4.101.3
 Path info:
    neighbor: Local           neighbor router id: 10.4.0.1
    valid local best
  Attributes after inbound policy was applied:
    next hop: 0.0.0.0
    MET ORG AS
    origin: IGP metric: 0
    aspath:
  Attributes after outbound policy was applied:
    next hop: 10.4.0.1
    MET ORG AS
    origin: IGP metric: 0
    aspath: 

12.0.0.0/24 is advertised to 10.4.101.2
 Path info:
    neighbor: 10.4.101.1       neighbor router id: 10.4.101.1
    valid external best
  Attributes after inbound policy was applied:
    next hop: 10.4.101.1
    ORG AS
    origin: IGP neighbor as: 2
    aspath: 2 3 4
  Attributes after outbound policy was applied:
    next hop: 10.4.101.1
    ORG AS
    origin: IGP neighbor as: 2
    aspath:2 3 4

12.0.0.0/24 is advertised to 10.4.101.3
 Path info:
    neighbor: 10.4.101.1       neighbor router id: 10.4.101.1
    valid external best
  Attributes after inbound policy was applied:
    next hop: 10.4.101.1
    ORG AS
    origin: IGP neighbor as: 2
    aspath: 2 3 4
  Attributes after outbound policy was applied:
    next hop: 10.4.101.1
    ORG AS
    origin: IGP neighbor as: 2
    aspath:2 3 4

This table describes the significant fields shown in the display.

Table 27. show bgp policy Field Descriptions

Field

Description

Is advertised to

IP address of the peer to which this route is advertised. If the route is advertised to multiple peers, information is shown separately for each peer.

neighbor

IP address of the peer that advertised this route, or one of the following:

Local—Route originated on the local system.

Local Aggregate—Route is an aggregate created on the local system.

neighbor router id

BGP identifier for the peer, or the local system if the route originated on the local system.

Not advertised to any peer

Indicates the no-advertise well-known community is associated with this route. Routes with this community are not advertised to any BGP peers.

Not advertised to any EBGP peer

Indicates the no-export well-known community is associated with this route. Routes with this community are not advertised to external BGP peers, even if those peers are in the same confederation as the local router.

Not advertised outside the local AS

Indicates the local-AS well-known community is associated with this route. Routes with this community value are not advertised outside the local autonomous system or confederation boundary.

(Received from a RR-client)

Path was received from a route reflector client.

(received-only)

Path is not used for routing purposes. It is used to support soft reconfiguration, and records the path attributes before inbound policy was applied to a path received from a peer. A path marked “received-only” indicates that either the path was dropped by inbound policy, or that a copy of path information was created and then modified for routing use.

(received & used)

Indicates that the path is used both for soft reconfiguration and routing purposes. A path marked “(received & used)”, implies the path information was not modified by inbound policy.

valid

Path is valid.

redistributed

Path is locally sourced through redistribution.

aggregated

Path is locally sourced through aggregation.

local

Path is locally sourced through the network command.

confed

Path was received from a confederation peer.

best

Path is selected as best.

multipath

Path is one of multiple paths selected for load-sharing purposes.

dampinfo

Indicates dampening information:

Penalty—Current penalty for this path.

Flapped—Number of times the route has flapped.

In—Time (hours:minutes:seconds) since the network first flapped.

Reuse in—Time (hours:minutes:seconds) after which the path is available. This field is displayed only if the path is currently suppressed.

Attributes after inbound policy was applied

Displays attributes associated with the received route, after any inbound policy has been applied.

AGG—Aggregator attribute is present.

AS—AS path attribute is present.

ATOM—Atomic aggregate attribute is present.

COMM—Communities attribute is present.

EXTCOMM—Extended communities attribute is present.

LOCAL—Local preference attribute is present.

MET—Multi Exit Discriminator (MED) attribute is present.

next hop—IP address of the next system used when a packet is forwarded to the destination network. An entry of 0.0.0.0 indicates that the router has a non-BGP route to this network.

ORG—Origin attribute is present.

origin

Origin of the path:

IGP—Path originated from an Interior Gateway Protocol (IGP) and was sourced by BGP using a network or aggregate-address command.

EGP—Path originated from an Exterior Gateway Protocol.

incomplete—Origin of the path is not clear; in example, a route that is redistributed into BGP from an IGP.

neighbor as

First autonomous system (AS) number in the AS path.

aggregator

Indicates that the path was received with the aggregator attribute. The AS number and router-id of the system that performed the aggregation are shown.

metric

Value of the interautonomous system metric, otherwise known as the MED metric.

localpref

Local preference value. This is used to determine the preferred exit point from the local autonomous system. It is propagated throughout the local autonomous system

aspath

AS path associated with the route.

community

Community attributes associated with the path. Community values are displayed in AA:NN format, except for the following well-known communities:

Local-AS—Community with value 4294967043 or hex 0xFFFFFF03. Routes with this community value are not advertised outside the local autonomous system or confederation boundary.

no-advertise—Community with value 4294967042 or hex 0xFFFFFF02. Routes with this community value are not advertised to any BGP peers.

no-export—Community with value 4294967041 or hex 0xFFFFFF01. Routes with this community are not advertised to external BGP peers, even if those peers are in the same confederation as the local router.

Extended community

Extended community attributes associated with the path. For known extended community types, the following codes may be displayed:

RT—Route target community

SoO—Site of Origin community

LB—Link Bandwidth community

Originator

Router ID of the originating router when route reflection is used.

Cluster lists

Router ID or cluster ID of all route reflectors through which the route has passed.

Attributes after outbound policy was applied

Displays attributes associated with the received route, after any outbound policy has been applied.

AGG—Aggregator attribute is present.

AS—AS path attribute is present.

ATOM—Atomic aggregate attribute is present.

COMM—Communities attribute is present.

EXTCOMM—Extended communities attribute is present.

LOCAL—Local preference attribute is present.

MET—Multi Exit Discriminator (MED) attribute is present.

next hop—IP address of the next system used when a packet is forwarded to the destination network. An entry of 0.0.0.0 indicates that the router has a non-BGP route to this network.

ORG—Origin attribute is present.

show bgp process

To display Border Gateway Protocol (BGP) process information, use the show bgp process command in EXEC mode.

show bgp [ipv4 | {unicast | multicast | labeled-unicast | all | tunnel | mdt} | ipv6 | {unicast | multicast | all | labeled-unicast} | all | {unicast | multicast | all | labeled-unicast | mdt | tunnel} | vpnv4 unicast | vpvn6 unicast] process [performance-statistics] [detail]

Syntax Description

ipv4

(Optional) Specifies IP Version 4.

unicast

(Optional) Specifies the unicast subaddress family.

multicast

(Optional) Specifies the multicast subaddress family.

labeled-unicast

(Optional) Specifies labeled unicast address prefixes.

all

(Optional) For subaddress families, specifies prefixes for all subaddress families.

tunnel

(Optional) Specifies tunnel address prefixes.

ipv6

(Optional) Specifies IP Version 6.

all

(Optional) For address family, specifies prefixes for all address families.

vpnv4 unicast

(Optional) Specifies VPNv4 unicast address families.

performance- statistics

(Optional) Displays performance statistics relative to the work done by the specified process.

detail

(Optional) Specifies detailed process information.

Command Default

If no address family or subaddress family is specified, the default address family and subaddress family specified using the set default-afi and set default-safi commands are used.

Command Modes

EXEC

Command History

Release

Modification

Release 3.7.2

This command was introduced.

Release 3.9.0

Asplain format for 4-byte autonomous system numbers notation was supported. The input parameters and output were modified to display 4-byte autonomous system numbers and extended communities in either asplain or asdot notations.

Release 5.3.2

The command output was modified to include graceful maintenance feature information.

Release 7.5.3

Updated output to display the delay of the BGP start-up process since the last router update.

Usage Guidelines

To use this command, you must be in a user group associated with a task group that includes appropriate task IDs. If the user group assignment is preventing you from using a command, contact your AAA administrator for assistance.


Note


The set default-afi command is used to specify the default address family for the session, and the set default-safi command is used to specify the default subaddress family for the session. See the System Management Command Reference for Cisco ASR 9000 Series Routers for detailed information and syntax for the set default-afi and set default-safi commands. If you do not specify a default address family, the default address family is IPv4. If you do not specify a default subaddress family, the default subaddress family is unicast.


Use the show bgp process command to display status and summary information for the Border Gateway Protocol (BGP) process. The output shows various global and address family-specific BGP configurations. A summary of the number of neighbors, update messages, and notification messages sent and received by the process is also displayed.

Use the detail keyword to display detailed process information. The detailed process information shows the memory used by each of various internal structure types.

Use the performance-statistics keyword to display a summary or detail of work done by the BGP processes. The summary display shows the real time spent performing certain operations and the time stamps for state transitions during initial convergence.

Task ID

Task ID

Operations

bgp

read

Examples

The following is sample output from the show bgp process command:



RP/0/RSP0/CPU0:router# show bgp process

BGP Process Information
BGP is operating in STANDALONE mode
Autonomous System: 1
Router ID: 10.0.0.5 (manually configured)
Cluster ID: 10.0.0.5
Fast external fallover enabled
Neighbor logging is enabled
Enforce first AS enabled
Default local preference: 100
Default keepalive: 60
Update delay: 120
Generic scan interval: 60

Address family: IPv4 Unicast
Dampening is enabled
Client reflection is enabled
Scan interval: 60
Main Table Version: 150
IGP notification: IGPs notified

Node         Process     Nbrs Estab Rst Upd-Rcvd Upd-Sent Nfn-Rcvd Nfn-Sent
node0_0_CPU0 Speaker        3     2   1       20       10        0        0

This table describes the significant fields shown in the display.

Table 28. show bgp process Field Descriptions

Field

Description

BGP is operating in

Indicates BGP is operating in standalone mode. This is the only supported mode.

Autonomous System

Autonomous system number for the local system.

  • Range for 2-byte Autonomous system numbers (ASNs) is 1 to 65535.
  • Range for 4-byte Autonomous system numbers (ASNs) in asplain format is 1 to 4294967295.
  • Range for 4-byte Autonomous system numbers (ASNs) is asdot format is 1.0 to 65535.65535.

Router ID

BGP identifier assigned to the local system. If this is explicitly configured using the bgp router-id command, “manually configured” is displayed. If the router ID is not manually configured, it is determined from a global router ID. If no global ID is available, the router ID is shown as 0.0.0.0.

Confederation ID

Confederation identifier for the local system.

Cluster ID

Cluster identifier for the local system. If this is manually configured using the bgp cluster-id command, “manually configured” is displayed.

Default metric

Default metric. This is controlled by the default-metric command.

Fast external fallover enabled

Indicates whether fast external fallover is enabled. This is controlled by the bgp fast-external-fallover disable command.

Neighbor logging enabled

Indicates whether logging of peer connection up and down transitions is enabled. This is controlled by the bgp log neighbor changes disable command.

Enforce first AS enabled

Indicates that strict checking of the first AS number in paths received from external BGP peers is enabled. This is controlled by the bgp enforce-first-as disable command.

iBGP to IGP redistribution

Indicates internal redistribution is enabled using the bgp redistribution-internal command.

Treating missing MED as worst

Indicates missing Multi Exit Discriminator (MED) metric values are treated as worst in the route selection algorithm. This is controlled by the bgp bestpath med missing-as-worst command.

Always compare MED is enabled

Indicates that the MED is always used during the route selection algorithm, even when paths are received from external BGP neighbors in different autonomous systems. This is controlled by the bgp bestpath med always command.

AS Path ignore is enabled

Indicates that the AS path length is ignored by the route selection algorithm. This is controlled by the bgp bestpath as-path ignore command.

Comparing MED from confederation peers

Indicates that the MED values are used in the route selection algorithm when comparing routes received from confederation peers. This is controlled by the bgp bestpath med confed command.

Comparing router ID for eBGP paths

Indicates that the router ID is used as a tiebreaker by the route selection algorithm when comparing identical routes received from different external BGP neighbors. This is controlled by the bgp bestpath compare-routerid command.

Default local preference

Default local preference value used for BGP routes. This is controlled by the bgp default local-preference command.

Default keepalive

Default keepalive interval. This is controlled by the timers bgp command.

Graceful restart enabled

Indicates that the graceful restart capability is enabled. The configuration commands affecting graceful restart behavior are:

  • bgp graceful-restart
  • bgp graceful-restart purge-time
  • bgp graceful-restart stalepath-time
  • bgp graceful-restart restart-time
  • bgp graceful-restart graceful-reset

Update delay

Maximum time that a BGP process stays in read-only mode.

Generic scan interval

Interval (in seconds) between BGP scans for address family-independent tasks. This is controlled by the bgp scan-time command.

Dampening

Indicates whether dampening is enabled for the specified address family. This is controlled by the dampening command.

Client reflection

Indicates whether client-to-client route reflection is enabled for the specified address family. This is controlled by the bgp client-to-client reflection disable command.

Scan interval

Interval (in seconds) between BGP scans for the given address family. This is controlled by the bgp scan-time command in address family configuration mode.

Main Table Version

Last version of the BGP database that was installed into the main routing table.

IGP notification

Indicates whether Interior Gateway Protocols (IGP) have been notified of BGP convergence for the specified address family.

Node

Node on which the process is executing.

Process

Type of BGP process.

Speaker

Speaker process. A speaker process is responsible for receiving, processing, and sending BGP messages to configured neighbors.

Nbrs

Number of neighbors for which the process is responsible.

Estab

Number of neighbors that have connections in the established state for this process.

Rst

Number of times this process was restarted.

Upd-Rcvd

Number of update messages received by the process.

Upd-Sent

Number of update messages sent by the process.

Nfn-Rcvd

Number of notification messages received by the process.

Nfn-Sent

Number of notification messages sent by the process.

The following is sample output from the show bgp process command with the Graceful Maintenance feature enabled:

RP/0/0/CPU0:R1#show bgp process
...
Graceful Maintenance active. Retaining routes in RIB during BGP shutdown
...

Or

Graceful Maintenance active for all neighbors. Retaining routes in RIB during BGP shutdown

**************************************************************************

RP/0/0/CPU0:Jan 28 22:01:36.356 : bgp[1056]: %ROUTING-BGP-5-ADJCHANGE : neighbor 10.10.10.4 Up (VRF: default) (AS: 4) WARNING: Graceful Maintenance is Active

The following is sample output from the show bgp process command with the detail keyword:


RP/0/RSP0/CPU0:router# show bgp all all process detail 

BGP Process Information
BGP is operating in STANDALONE mode
Autonomous System: 1
Router ID: 10.0.0.5 (manually configured)
Cluster ID: 10.0.0.5
Fast external fallover enabled
Neighbor logging is enabled
Enforce first AS enabled
Default local preference: 100
Default keepalive: 60
Update delay: 120
Generic scan interval: 60

BGP Speaker process: 0, location node0_0_0
Neighbors: 3, established: 2

                           Sent            Received
Updates:                   3               15              
Notifications:             0               0               

                           Number          Memory Used
Attributes:                12              1104            
AS Paths:                  10              400             
Communities:               2               1080            
Extended communities:      1               40              
Route Reflector Entries:   0               0               
Route-map Cache Entries:   0               0               
Filter-list Cache Entries: 0               0               
Next Hop Cache Entries:    2               80              
Update messages queued:    0               

Address family: IPv4 Unicast
Dampening is enabled
Client reflection is enabled
Main Table Version: 12
IGP notification: IGPs notified

State: normal mode.
BGP Table Version: 12
Network Entries: 15, Soft Reconfig Entries: 0
Dampened Paths: 0, History Paths: 9

                           Allocated       Freed
Prefixes:                  15              0               
Paths:                     19              0               

                           Number          Memory Used
Prefixes:                  15              1230            
Paths:                     19              760 

This table describes the significant fields shown in the display.

Table 29. show bgp process detail Field Descriptions

Field

Description

BGP is operating in

Indicates whether BGP is operating in standalone mode.

Autonomous System

Autonomous system number for the local system.

Router ID

BGP identifier assigned to the local system. If this is explicitly configured using the bgp router-id command, “manually configured” is displayed. If the router ID is not manually configured, it is determined from a global router ID. If the global ID is not available, the router ID is shown as 0.0.0.0.

Confederation ID

Confederation identifier for the local system.

Cluster ID

Cluster identifier for the local system. If this is manually configured using the bgp cluster-id command, “manually configured” is displayed.

Default metric

Default metric.

Fast external fallover enabled

Indicates whether fast external fallover is enabled.

Neighbor logging enabled

Indicates whether logging of peer connection up and down transitions is enabled.

Enforce first AS enabled

Indicates that strict checking of the first autonomous system (AS) number in paths received from external BGP peers is enabled.

iBGP to IGP redistribution

Indicates internal redistribution is enabled using the bgp redistribution-internal command.

Treating missing MED as worst

Indicates missing MED metric values are treated as worst in the route selection algorithm. This is controlled by the bgp bestpath med missing-as-worst command.

Always compare MED is enabled

Indicates that the MED is always used during the route selection algorithm, even when paths are received from external BGP neighbors in different autonomous systems. This is controlled by the bgp bestpath med always command.

AS Path ignore is enabled

Indicates that the AS path length is ignored by the route selection algorithm. This is controlled by the bgp bestpath as-path ignore command.

Comparing MED from confederation peers

Indicates that the MED values are used in the route selection algorithm when comparing routes received from confederation peers. This is controlled by the bgp bestpath med confed command.

Comparing router ID for eBGP paths

Indicates that the router ID is used as a tiebreaker by the route selection algorithm when comparing identical routes received from different external BGP neighbors. This is controlled by the bgp bestpath compare-routerid command.

Default local preference

Default local preference value used for BGP routes.

Default keepalive

Default keepalive interval. This is controlled by the timers bgp command.

Graceful restart enabled

Indicates that the graceful restart capability is enabled. The configuration commands affecting graceful restart behavior are:

  • bgp graceful-restart
  • bgp graceful-restart purge-time
  • bgp graceful-restart stalepath-time
  • bgp graceful-restart restart-time
  • bgp graceful-restart graceful-reset

Update delay

Maximum time that a BGP process stays in read-only mode.

Generic scan interval

Interval (in seconds) between BGP scans for address family-independent tasks. This is controlled by the bgp scan-time command.

BGP Speaker Process

Speaker process responsible for receiving, processing and sending BGP messages.

Node

Node on which the specified process is executing.

Neighbors

Number of neighbors for which the specified process is responsible.

established

Number of neighbors that have connections in the established state for the specified process.

Updates

Number of update messages sent and received by the specified process.

Notifications

Number of notification messages sent and received by the specified process.

Attributes

Number of unique sets of attribute information stored in the specified process and the amount of memory used by the attribute information.

AS Paths

Number of unique autonomous system paths stored in the specified process and the amount of memory used by the AS path information.

Communities

Number of unique sets of community information stored in the specified process and the amount of memory used by them.

Extended communities

Number of unique sets of extended community information stored in the specified process and the amount of memory used by them.

Route Reflector Entries

Number of unique sets of route reflector information stored in the specified process and the amount of memory used by them.

Nexthop Entries

Number of entries and memory usage for cached next- hop information.

Update messages queued

Total number of update messages queued to be sent across all neighbors for which the specified process is responsible.

Address family

Specified address family.

Dampening

Indicates whether dampening is enabled for the specified address family.

Client reflection

Indicates whether client-to-client route reflection is enabled for the specified address family. This is controlled by the bgp client-to-client reflection disable command.

Scan interval

Interval (in seconds) between BGP scans for the given address family. This is controlled by the bgp scan-time command.

Main Table Version

Last version of the local BGP database for the specified address family that was injected into the main routing table.

IGP notification

Indicates whether IGPs have been notified of BGP convergence for the specified address family.

RIB has converged

Indicates whether the main routing table version has converged and the version at which it converged.

State

BGP system state for the specified address family and process. This may be one of the following:

read-only mode—Initial set of updates is being recovered. In this mode, route selection is not performed, routes are not installed in the global RIB, and updates are not advertised to peers.

best-path calculation mode—Route selection is being performed for the routes that were received while in read-only mode.

import mode—Routes are imported from one VRF to another VRF once the best paths are calculated. This mode is supported in VPNv4 unicast address family mode.

RIB update mode—Routes that were selected in best-path calculation mode are being installed in the global RIB.

label allocation mode: Labels are allocated for the received prefixes based on the requirement.

normal mode—Best paths are sent to the peers for routes that exist in the RIB. The route selection, import processing, RIB updates, and label allocation are performed as new updates are received.

BGP Table Version

Last version used in the BGP database for received routes.

Attribute download

Indicates whether the RIB attribute download is enabled.

Network Entries

Number of sets of prefix information held in the specified BGP process for the specified address family.

Soft Reconfig Entries

Number of sets of prefix information that are present only for the purpose of supporting soft reconfiguration.

Dampened Paths

Number of routes that are suppressed due to dampening for the specified address family.

History Paths

Number of routes that are currently withdrawn, but are being maintained to preserve dampening information.

Prefixes (Allocated/Freed)

Number of sets of prefix information for the specified address family that have been allocated and freed during the lifetime of the process.

Paths (Allocated/Freed)

Number of sets of route information for the specified address family that have been allocated and freed during the lifetime of the process.

Prefixes (Number/Memory Used)

Number of sets of prefix information currently allocated for the specified address family, and the amount of memory used by them.

Paths (Number/Memory Used)

Number of sets of route information currently allocated for the specified address family, and the amount of memory used by them.

The following is sample output from the show bgp process command with the performance-statistics keyword:


RP/0/RSP0/CPU0:router# show bgp process performance-statistics detail

BGP Speaker process: 0, Node: node0_0_CPU0
Restart count: 2
Neighbors: 3, established: 2

                           Sent            Received
Updates:                   20              20              
Notifications:             0               0               

                           Number          Memory Used
Attributes:                2               184             
AS Paths:                  2               48              
Communities:               0               0               
Extended communities:      0               0               
Route Reflector Entries:   0               0               
Route-map Cache Entries:   0               0               
Filter-list Cache Entries: 0               0               
Next Hop Cache Entries:    2               80              
Update messages queued:    0               

Read 14 messages (1142 bytes) in 12 calls (time spent: 0.024 secs)
Read throttled 0 times
Processed 14 inbound messages (time spent: 0.132 secs)
Wrote 2186 bytes in 24 calls (time spent: 0.024 secs)
Processing write list: wrote 18 messages in 4 calls (time spent: 0.000 secs)
Processing write queue: wrote 10 messages in 20 calls (time spent: 0.000 secs)
Socket setup (LPTS): 4 calls (time spent: 0.010 secs)
Configuration: 1 requests (time spent: 0.002 secs)
Operational data: 9 requests (time spent: 0.026 secs)


State: normal mode.
BGP Table Version: 150
Network Entries: 149, Soft Reconfig Entries: 0



                           Allocated       Freed
Prefixes:                  149             0               
Paths:                     200             0               

                           Number          Memory Used
Prefixes:                  149             12516           
Paths:                     200             8000            

Updates generated: 149 prefixes in 8 messages from 2 calls (time spent: 0.046 secs)
Scanner: 2 scanner runs (time spent: 0.008 secs)
RIB update: 1 rib update runs, 149 prefixes installed (time spent: 0.024 secs)
Process has converged for IPv4 Unicast.




First neighbor established: 1082604050s
Entered DO_BESTPATH mode: 1082604055s
Entered DO_RIBUPD mode: 1082604055s
Entered Normal mode: 1082604055s
Latest UPDATE sent: 1082604056s

This table describes the significant fields shown in the display.

Table 30. show bgp process performance-statistics Field Descriptions

Field

Description

BGP is operating in

Indicates whether BGP is operating in standalone mode.

Autonomous system

Autonomous system number for the local system.

Router ID

BGP identifier assigned to the local system. If this is explicitly configured using the bgp router-id command, “manually configured” is displayed. If the router ID is not manually configured, it is determined from a global router ID. If the global ID is not available, the router ID is shown as 0.0.0.0.

Confederation ID

Confederation identifier for the local system.

Cluster ID

The cluster identifier for the local system. If this is manually configured using the bgp cluster-id command, “manually configured” is displayed.

Default metric

Default metric.

Fast external fallover enabled

Indicates whether fast external fallover is enabled.

Neighbor logging enabled

Indicates whether logging of peer connection up and down transitions is enabled. This is controlled by the bgp log neighbor changes disable command.

Enforce first AS enabled

Indicates that strict checking of the first AS number in paths received from external BGP peers is enabled.

iBGP to IGP redistribution

Indicates internal redistribution is enabled using the bgp redistribution-internal command.

Treating missing MED as worst

Indicates missing MED metric values are treated as worst in the route selection algorithm. This is controlled using the bgp bestpath med missing-as-worst command.

Always compare MED is enabled

Indicates that the MED is always used during the route selection algorithm, even when paths are received from external BGP neighbors in different autonomous systems. This setting is controlled by the bgp bestpath med always command.

AS Path ignore is enabled

Indicates that the AS path length is ignored by the route selection algorithm. This is controlled by the bgp bestpath as-path ignore command.

Comparing MED from confederation peers

Indicates that the MED values are used in the route selection algorithm when comparing routes received from confederation peers. This is controlled by the bgp bestpath med confed command.

Comparing router ID for eBGP paths

Indicates that the router ID is used as a tiebreaker by the route selection algorithm when comparing identical routes received from different external BGP neighbors. This is controlled by the bgp bestpath compare-routerid command.

Default local preference

Default local preference value used for BGP routes.

Default keepalive

Default keepalive interval. This setting is controlled by the timers bgp command.

Graceful restart enabled

Indicates that the graceful restart capability is enabled. The configuration commands affecting graceful restart behavior are: bgp graceful-restart, bgp graceful-restart purge-time, bgp graceful-restart stalepath-time, bgp graceful-restart restart-time, and bgp graceful-restart graceful-reset.

Update delay

Maximum time that a BGP process stays in read-only mode.

Generic scan interval

Interval (in seconds) between BGP scans for address family-independent tasks. This setting is controlled by the bgp scan-time command in router configuration mode.

Address family

Specified address family.

Dampening

Indicates whether dampening is enabled for the specified address family.

Client reflection

Indicates whether client-to-client route reflection is enabled for the specified address family. This is controlled by the bgp client-to-client reflection disable command.

Scan interval

Interval (in seconds) between BGP scans for the given address family. This is controlled by the bgp scan-time command.

Main Table Version

Last version of the local BGP database for the specified address family that was injected into the main routing table.

IGP notification

Indicates whether IGPs have been notified of BGP convergence for the specified address family.

Node

Node on which the process is executing.

Process

BGP process.

Speaker

Speaker process. The speaker process is responsible for receiving, processing and sending BGP messages.

Read

Real time (in seconds) spent reading messages from peers by this process.

Write

Real time (in seconds) spent writing messages to peers by this process.

Inbound

The real time (in seconds) spent processing messages read from peers by this process.

Config

Real time (in seconds) spent processing configuration commands by this process.

Data

Real time (in seconds) spent providing operational data by this process.

Conv

Indicates whether the process has converged after the initial update.

Nbr Estab

Time stamp (in seconds) recording the time when the first neighbor became established.

Bestpath

Time stamp (in seconds) recording the time the best-path calculation mode was entered.

RIB Inst

Time stamp (in seconds) recording the time RIB update mode was entered.

Read/Write

Time stamp (in seconds) recording the time normal mode was entered.

Last Upd

Time stamp (in seconds) recording the time the last update was sent to a neighbor.

Address Family IPv4 Unicast converged in n seconds

Indicates that BGP has reached initial convergence for the IPv4 unicast address family. The time taken for convergence is shown.

Address Family IPv6 Multicast converged in n seconds

Indicates that BGP has reached initial convergence for the IPv6 multicast address family. The time taken for convergence is shown.

The following is sample output from the show bgp process command with the performance-statistics and detail keywords:


RP/0/RSP0/CPU0:router# show bgp process performance-statistics detail

BGP Speaker process: 0, Node: node0_0_CPU0
Restart count: 2
Neighbors: 3, established: 2

                           Sent            Received
Updates:                   20              20              
Notifications:             0               0               

                           Number          Memory Used
Attributes:                2               184             
AS Paths:                  2               48              
Communities:               0               0               
Extended communities:      0               0               
Route Reflector Entries:   0               0               
Route-map Cache Entries:   0               0               
Filter-list Cache Entries: 0               0               
Next Hop Cache Entries:    2               80              
Update messages queued:    0               

Read 14 messages (1142 bytes) in 12 calls (time spent: 0.024 secs)
Read throttled 0 times
Processed 14 inbound messages (time spent: 0.132 secs)
Wrote 2186 bytes in 24 calls (time spent: 0.024 secs)
Processing write list: wrote 18 messages in 4 calls (time spent: 0.000 secs)
Processing write queue: wrote 10 messages in 20 calls (time spent: 0.000 secs)
Socket setup (LPTS): 4 calls (time spent: 0.010 secs)
Configuration: 1 requests (time spent: 0.002 secs)
Operational data: 9 requests (time spent: 0.026 secs)


State: normal mode.
BGP Table Version: 150
Network Entries: 149, Soft Reconfig Entries: 0



                           Allocated       Freed
Prefixes:                  149             0               
Paths:                     200             0               

                           Number          Memory Used
Prefixes:                  149             12516           
Paths:                     200             8000            

Updates generated: 149 prefixes in 8 messages from 2 calls (time spent: 0.046 secs)
Scanner: 2 scanner runs (time spent: 0.008 secs)
RIB update: 1 rib update runs, 149 prefixes installed (time spent: 0.024 secs)
Process has converged for IPv4 Unicast.

First neighbor established: 1082604050s
Entered DO_BESTPATH mode: 1082604055s
Entered DO_RIBUPD mode: 1082604055s
Entered Normal mode: 1082604055s
Latest UPDATE sent: 1082604056s

This table describes the significant fields shown in the display.

Table 31. show bgp process performance-statistics detail Field Descriptions

Field

Description

Process

The specified process.

Location

Node in which the specified process is executing.

Neighbors

Number of neighbors for which the specified process is responsible.

established

Number of neighbors that have connections in the established state for the specified process.

Updates

Number of update messages sent and received by the specified process.

Notifications

Number of notification messages sent and received by the specified process.

Attributes

Number of unique sets of attribute information stored in the specified process and the amount of memory used by the attribute information.

AS Paths

Number of unique autonomous system paths stored in the specified process, and the amount of memory used by the AS path information.

Communities

Number of unique sets of community information stored in the specified process and the amount of memory used by them.

Extended communities

Number of unique sets of extended community information stored in the specified process and the amount of memory used by them.

Route Reflector Entries

Number of unique sets of route reflector information stored in the specified process and the amount of memory used by them.

Route-map Cache Entries

Number of entries and memory usage for cached results for applying a route map.

Filter-list Cache Entries

Number of entries and memory usage for cached results for applying an AS path filter list.

Next Hop Cache Entries

Number of entries and memory usage for cached next-hop information.

Update messages queued

Number of update messages queued to be sent across all neighbors for which the specified process is responsible.

Read

Indicates the number of messages read by the process, the total size of read messages, the number of read operations performed, and the real time spent by the process performing read operations.

Read throttled

Number of times that reading from TCP has been throttled due to a backlog of messages read but not processed.

inbound messages

Number of read messages that have been processed and the real time spent processing inbound messages.

Wrote

Amount of data that has been written by the process, the number of write operations performed, and the real time spent by the process performing write operations.

Processing write list

Number of messages written from write lists, the number of times the write list has been processed, and the real time spent processing the write list.

Note

 

Write lists typically contain only update messages.

Processing write queue

Number of messages written from write queues, number of times the write queue has been processed, and the real time spent processing the write queue.

Socket setup

Number of socket setup operations performed and the real time spent during socket setup operations.

Configuration

Number of configuration requests received by the process and the real time spent processing configuration requests.

Operational data

Number of requests for operational data (for show commands) received by the process and the real time spent processing operation data requests

State

BGP system state for the specified address family and process. This may be one of the following:

read-only mode—Initial set of updates is being recovered. In this mode, route selection is not performed, routes are not installed in the global RIB, and updates are not advertised to peers.

best-path calculation mode—Route selection is being performed for the routes that were received while in read-only mode.

import mode—Routes are imported from one VRF to another VRF once the best paths are calculated. This mode is supported in VPNv4 unicast address family mode.

RIB update mode—Routes that were selected in best-path calculation mode are being installed in the global RIB.

label allocation mode: Labels are allocated for the received prefixes based on the requirement.

normal mode—Best paths are sent to the peers for routes that exist in the RIB. The route selection, import processing, RIB updates, and label allocation are performed as new updates are received.

BGP Table Version

Last version used in the BGP database for received routes.

Network Entries

Number of sets of prefix information held in the specified BGP process for the specified address family.

Soft Reconfig Entries

Number of sets of prefix information that are present only for the purpose of supporting soft reconfiguration.

Dampened Paths

Number of routes that are suppressed due to dampening for the specified address family.

History Paths

Number of routes that are currently withdrawn, but are being maintained to preserve dampening information.

Prefixes (Allocated/Freed)

Number of sets of prefix information for the specified address family that have been allocated and freed during the lifetime of the process.

Paths (Allocated/Freed)

Number of sets of route information for the specified address family that have been allocated and freed during the lifetime of the process.

Prefixes (Number/Memory Used)

Number of sets of prefix information currently allocated for the specified address family and amount of memory used by them.

Paths (Number/Memory Used)

Number of sets of route information currently allocated for the specified address family and amount of memory used by them.

Updates generated

Number of prefixes for which updates have been generated, the number of messages used to advertise the updates, the number of update generation runs performed, and the real time spent generating updates for the specified address family.

Scanner

Number of times the scanner has run for the specified address family and real time spent in scanner processing.

RIB Update

Number of global routing information base update runs performed for the specified address family, number of prefixes installed, withdrawn, or modified in the global RIB during these runs, and real time spent performing these runs.

Process has converged

Indicates whether the process has reached initial convergence for the specified address family.

First neighbor established

Time stamp (in seconds) recording the time the first neighbor in the process was established.

Entered DO_BESTPATH mode

Time stamp (in seconds) recording the time best-path calculation mode was entered.

Entered DO_RIBUPD mode

Time stamp (in seconds) recording the time RIB update mode was entered.

Entered Normal mode

Time stamp (in seconds) recording the time normal mode was entered.

Last UPDATE sent

Time stamp (in seconds) recording the time the last update was sent to a neighbor.

The following is sample output from the show bgp vpnv4 unicast process performance-statistics detail command:

RP/0/RSP0/CPU0:router# show bgp vpnv4 unicast process performance-statistics detail
BGP Speaker process: 0, Node: node0_8_CPU0 Restart count: 1
                           Total           Nbrs Estab/Cfg
Default VRFs:              1               4/12
Non-Default VRFs:          1009            1082/1337

                           Sent            Received
Updates:                   362259          5688505         
Notifications:             14              0               

                           Number          Memory Used
Attributes:                14896           2979200         
AS Paths:                  17              1100            
Communities:               3               120             
Extended communities:      1849            124440          
Route Reflector Entries:   417             25020           
Nexthop Entries:           2941            539572          
Update messages queued:    0               

                           Alloc           Free          
Pool 210:                  28955629        28955628      
Pool 310:                  363103          363103        
Pool 600:                  4931162         4931162       
Pool 1100:                 104693          104693        
Pool 4300:                 799374          799374        

Read 34755745 messages (3542094326 bytes) in 30528983 calls (time spent: 6427.769 secs) Read partly throttled 1506 times
    Read 14 times after crossing lower threshold Processed 5836892 inbound update messages (time spent: 6229.512 secs)
	Wrote 825719955 bytes in 29272669 calls (time spent: 2318.472 secs) 
	Processing sub-group: wrote 861402 messages in 1113810 calls (time spent: 145.446 secs)
	Processing write queue: wrote 6288 messages in 20498 calls (time spent: 0.039 secs) 
	Socket setup (LPTS): 0 calls (time spent: 0.000 secs) 
	event_file_attach calls: Input 8769, Output 2810, Input-output 0
Configuration: 989 requests (time spent: 0.046 secs) Operational data: 92396 requests (time spent: 98.864 secs) 
Current Clock Time: not set Update Generation master timer: 
    id: 0, time left: 0.0 sec, last processed: not set
    expiry time of parent node: not set
IO master timer: 
    id: 0, time left: 0.0 sec, last processed: not set
    expiry time of parent node: not set


Address Family: VPNv4 Unicast
State: Normal mode.
BGP Table Version: 23211188
Attribute download: Disabled
Soft Reconfig Entries: 0

                   Last 8 Triggers       Ver         Tbl Ver   

Label Thread       Jun 18 05:31:39.120   23211188    23211188  
                   Jun 18 05:31:35.274   23211188    23211188  
                   Jun 18 05:31:34.340   23211187    23211188  
                   Jun 18 05:31:34.189   23211186    23211187  
                   Jun 18 05:31:29.120   23211186    23211186  
                   Jun 18 05:31:28.861   23211186    23211186  
                   Jun 18 05:31:19.640   23211186    23211186  
                   Jun 18 05:31:19.272   23211186    23211186  
                   Total triggers: 639526

Import Thread      Jun 18 05:31:39.120   23211188    23211188  
                   Jun 18 05:31:35.274   23211188    23211188  
                   Jun 18 05:31:34.340   23211187    23211188  
                   Jun 18 05:31:34.189   23211186    23211187  
                   Jun 18 05:31:29.120   23211186    23211186  
                   Jun 18 05:31:28.861   23211186    23211186  
                   Jun 18 05:31:19.640   23211186    23211186  
                   Jun 18 05:31:19.272   23211186    23211186  
                   Total triggers: 689177

RIB Thread         Jun 18 05:31:39.146   23211188    23211188  
                   Jun 18 05:31:35.299   23211188    23211188  
                   Jun 18 05:31:34.525   23211187    23211188  
                   Jun 18 05:31:34.494   23211186    23211188  
                   Jun 18 05:31:34.340   23211186    23211188  
                   Jun 18 05:31:34.255   23211186    23211188  
                   Jun 18 05:31:29.146   23211186    23211186  
                   Jun 18 05:31:28.886   23211186    23211186  
                   Total triggers: 668084

Update Thread      Jun 18 05:31:39.171   ---         23211188  
                   Jun 18 05:31:35.324   ---         23211188  
                   Jun 18 05:31:34.558   ---         23211188  
                   Jun 18 05:31:34.521   ---         23211188  
                   Jun 18 05:31:34.327   ---         23211188  
                   Jun 18 05:31:29.170   ---         23211186  
                   Jun 18 05:31:28.910   ---         23211186  
                   Jun 18 05:31:19.690   ---         23211186  
                   Total triggers: 660143

                      Allocated       Freed         
Remote Prefixes:      3150972         2885064       
Remote Paths:         7639074         7118286       

Local Prefixes:       3760870         3425614       
Local Paths:          7892100         7595657       

                      Number          Mem Used      
Remote Prefixes:      265908          29781696      
Remote Paths:         520788          24997824      
Remote RDs:           12424           2832672       

Local Prefixes:       335256          37548672      
Local Paths:          296443          14229264      
Local RDs:            1009            230052        

Total Prefixes:       601164          67330368      
Total Paths:          817231          39227088      
Imported Paths:       265675          12752400      
Total RDs:            13433           3062724       
Same RDs:             0               0             


Update Groups: 3  Subgroups: 2
Updates generated: 1438448 prefixes in 67375 messages from 181564 calls (time spent: 6779.576 secs)
Scanner: 0 scanner runs (time spent: 0.000 secs) RIB update: 0 rib update runs, 0 prefixes installed, 0 modified,
            0 prefixes removed (time spent: 0.000 secs) RIB table update: 0 table deletes, 0 table invalid, 3526736604 table skip,
            0 no local label, 0 rib retries Process has not converged for VPNv4 Unicast.

First neighbor established:  Jun 11 08:32:10
Entered DO_BESTPATH mode:    Jun 11 08:52:10
Entered DO_IMPORT mode:      Jun 11 08:52:12
Entered DO_LABEL_ALLOC mode: Jun 11 08:52:16
Entered DO_RIBUPD mode:      Jun 11 08:52:19
Entered Normal mode:         Jun 11 08:52:23
Latest UPDATE sent:          Jun 18 05:31:34

The Update wait-install enabled section in the show bgp process command displays the delay of the BGP process update since the last router reload.
Router# show bgp process
Wed Aug 24 00:40:48.649 PDT

BGP Process Information:
BGP is operating in STANDALONE mode
Autonomous System number format: ASPLAIN
Autonomous System: 100
Router ID: 192.168.0.2 (manually configured)
Default Cluster ID: 192.168.0.2
Active Cluster IDs:  192.168.0.2
------------------------------
------------------------------
Update wait-install enabled:
  ack request 2, ack rcvd 2, slow ack 0
  startup delay 10 secs

--More—

show bgp regexp

To display routes matching the autonomous system path regular expression, use theshow bgp regexp command in EXECEXEC mode mode.

show bgp regexp regular-expression

Syntax Description

ipv4

(Optional) Specifies IP Version 4 address prefixes.

unicast

(Optional) Specifies unicast address prefixes.

multicast

(Optional) Specifies multicast address prefixes.

labeled-unicast

(Optional) Specifies labeled unicast address prefixes.

all

(Optional) For subaddress families, specifies prefixes for all subaddress families.

tunnel

(Optional) Specifies tunnel address prefixes.

ipv6

(Optional) Specifies IP Version 6 address prefixes.

all

(Optional) For address family, specifies prefixes for all address families.

vpnv4 unicast

(Optional) Specifies VPNv4 unicast address families.

vrf

(Optional) Specifies VPN routing and forwarding (VRF) instance.

vrf-name

(Optional) Name of a VRF.

all

(Optional) For VRF, specifies all VRFs.

ipv4 { unicast | labeled-unicast }

(Optional) For VRF, specifies IPv4 unicast or labeled-unicast address families.

ipv6 unicast

(Optional) For VRF, specifies IPv6 unicast address families.

regular-expression

Regular expression to match the BGP autonomous system paths.

Command Default

If no address family or subaddress family is specified, the default address family and subaddress family specified using the set default-afi and set default-safi commands are used.

Command Modes

EXECEXEC mode

Command History

Release

Modification

Release 3.7.2

This command was introduced.

Usage Guidelines

To use this command, you must be in a user group associated with a task group that includes appropriate task IDs. If the user group assignment is preventing you from using a command, contact your AAA administrator for assistance.


Note


The set default-afi command is used to specify the default address family for the session, and the set default-safi command is used to specify the default subaddress family for the session. See the System Management Command Reference for Cisco ASR 9000 Series Routers for detailed information and syntax for the set default-afi and set default-safi commands. If you do not specify a default address family, the default address family is IPv4. If you do not specify a default subaddress family, the default subaddress family is unicast.


BGP contains a separate routing table for each configured address family and subaddress family combination. The address family and subaddress family options specify the routing table to be examined. If the all keyword is specified for the address family or subaddress family, each matching routing table is examined in turn.

Use the show bgp regexp command to display all routes in the specified BGP table whose autonomous system path is matched by the specified regular expression.


Note


If the regular expression contains spaces and parentheses, it must be specified and surrounded by quotation marks.


Task ID

Task ID

Operations

bgp

read

Examples

The following is sample output from the show bgp regexp command:


Router# show bgp regexp "^3 "

BGP router identifier 10.0.0.5, local AS number 1
BGP main routing table version 64
BGP scan interval 60 secs
Status codes: s suppressed, d damped, h history, * valid, > best
              i - internal, S stale
Origin codes: i - IGP, e - EGP, ? - incomplete
   Network      Next Hop            Metric LocPrf Weight Path
*>i172.20.17.121      10.0.101.2                    100      0 3 2000 3000 i
*>i10.0.0.0           10.0.101.2                    100      0 3 100 1000 i
*>i172.5.23.0/24      10.0.101.2                    100      0 3 4 60 4378 i

This table describes the significant fields shown in the display.

Table 32. show bgp regexp Field Descriptions

Field

Description

BGP router identifier

BGP identifier for the local system.

local AS number

Autonomous system number for the local system.

Dampening enabled

Displayed if dampening has been enabled for the routes in this BGP routing table.

BGP main routing table version

Last version of the BGP database that was installed into the main routing table.

BGP scan interval

Interval (in seconds) between scans of the BGP table specified by the address family and subaddress family.

Status codes

Status of the table entry. The status is displayed as a three-character field at the beginning of each line in the table. The first character may be (in order of precedence):

S—Path is stale, indicating that a graceful restart is in progress with the peer from which the route was learned.

s—Path is more specific than a locally sourced aggregate route and has been suppressed.

*—Path is valid.

The second character may be (in order of precedence):

>—Path is the best path to use for that network.

d—Path is dampened.

h—Path is a history entry, representing a route that is currently withdrawn, but that is being maintained to preserve dampening information. Such routes should never be marked as valid.

The third character may be:

i—Path was learned by an internal BGP (iBGP) session.

Origin codes

Origin of the path. The origin code is placed at the end of each line in the table. It can be one of the following values:

i—Path originated from an Interior Gateway Protocol (IGP) and was advertised with a network or aggregate-address command.

e—Path originated from an Exterior Gateway Protocol (EGP).

?—Origin of the path is not clear. Usually, this is a route that is redistributed into BGP from an IGP.

Network

IP address of a network entity.

Next Hop

IP address of the next system that is used when a packet is forwarded to the destination network. An entry of 0.0.0.0 indicates that the router has a non-BGP route to this network.

Metric

Value of the interautonomous system metric, otherwise known as the Multi Exit Discriminator (MED) metric.

LocPrf

Local preference value. This is used to determine the preferred exit point from the local autonomous system. It is propagated throughout the local autonomous system.

Weight

Path weight. Weight is used in choosing the preferred path to a route. It is not advertised to any neighbor.

Path

Autonomous system path to the destination network. At the end of the path is the origin code for the path.

show bgp route-policy

To display Border Gateway Protocol (BGP) information about networks that match an outbound route policy, use the show bgp route-policy command in EXECEXEC mode mode.

show bgp route-policy route-policy-name []

Syntax Description

ipv4

(Optional) Specifies IP Version 4 address prefixes.

unicast

(Optional) Specifies unicast address prefixes.

multicast

(Optional) Specifies multicast address prefixes.

labeled-unicast

(Optional) Specifies labeled unicast address prefixes.

all

(Optional) For subaddress families, specifies prefixes for all subaddress families.

tunnel

(Optional) Specifies tunnel address prefixes.

ipv6

(Optional) Specifies IP Version 6 address prefixes.

all

(Optional) For address family, specifies prefixes for all address families.

vpnv4 unicast

(Optional) Specifies VPNv4 unicast address families.

rd rd-address

(Optional) Displays routes with a specific route distinguisher.

vrf

(Optional) Specifies VPN routing and forwarding (VRF) instance.

vrf-name

(Optional) Name of a VRF.

all

(Optional) For VRF, specifies all VRFs.

ipv4 { unicast | labeled-unicast }

(Optional) For VRF, specifies IPv4 unicast or labeled-unicast address families.

ipv6 unicast

(Optional) For VRF, specifies IPv6 unicast address families.

route-policy-name

Name of a route policy.

Command Default

If no address family or subaddress family is specified, the default address family and subaddress family specified using the set default-afi and set default-safi commands are used.

Command Modes

EXECEXEC mode

Command History

Release

Modification

Release 3.7.2

This command was introduced.

Usage Guidelines

To use this command, you must be in a user group associated with a task group that includes appropriate task IDs. If the user group assignment is preventing you from using a command, contact your AAA administrator for assistance.


Note


The set default-afi command is used to specify the default address family for the session, and the set default-safi command is used to specify the default subaddress family for the session. See the System Management Command Reference for Cisco ASR 9000 Series Routers for detailed information and syntax for the set default-afi and set default-safi commands. If you do not specify a default address family, the default address family is IPv4. If you do not specify a default subaddress family, the default subaddress family is unicast.


BGP contains a separate routing table for each address family and subaddress family combination that has been configured. The address family and subaddress family options specify the routing table to be examined. If the all keyword is specified for the address family or subaddress family, each matching routing table is examined.

A route policy must be configured to use this command. When the show bgp route-policy command is entered, routes in the specified BGP table are compared with the specified route policy, and all routes passed by the route policy are displayed.

If a pass clause is encountered while the route policy is being applied to the route and the route policy processing completes without hitting a drop clause, the route is displayed. The route is not displayed if a drop clause is encountered, if the route policy processing completes without hitting a pass clause, or if the specified route policy does not exist.

The information displayed does not reflect modifications the policy might make to the route. To display such modifications, use the show bgp policy command.

Task ID

Task ID

Operations

bgp

read

Examples

The following is sample output from the show bgp route-policy command in EXECEXEC mode mode:


Router# show bgp route-policy p1

BGP router identifier 172.20.1.1, local AS number 1820
BGP main routing table version 729
Dampening enabled
BGP scan interval 60 secs
Status codes: s suppressed, d damped, h history, * valid, > best
              i - internal, S stale
Origin codes: i - IGP, e - EGP, ? - incomplete
   Network            Next Hop            Metric LocPrf Weight Path
*  10.13.0.0/16       192.168.40.24                          0 1878 704 701 200 ?
*  10.16.0.0/16       192.168.40.24                          0 1878 704 701 i

This table describes the significant fields shown in the display.

Table 33. show bgp route-policy Field Descriptions

Field

Description

BGP router identifier

BGP identifier for the local system.

local AS number

Autonomous system number for the local system.

BGP main routing table version

Last version of the BGP database that was installed into the main routing table.

Dampening enabled

Displayed if dampening is enabled for the routes in this BGP routing table.

BGP scan interval

Interval (in seconds) between scans of the BGP table specified by the address family and subaddress family.

Status codes

Status of the table entry. The status is displayed as a three-character field at the beginning of each line in the table. The first character may be (in order of precedence):

S—Path is stale, indicating that a graceful restart is in progress with the peer from which the route was learned.

s—Path is more specific than a locally sourced aggregate route and has been suppressed.

*—Path is valid.

The second character may be (in order of precedence):

>—Path is the best path to use for that network.

d—Path is dampened.

h—Path is a history entry, representing a route that is currently withdrawn, but that is being maintained to preserve dampening information. Such routes should never be marked as valid.

The third character may be:

i—Path was learned by an internal BGP (iBGP) session.

Origin codes

Origin of the path. The origin code is displayed at the end of each line in the table. It can be one of the following values:

i—Path originated from an Interior Gateway Protocol (IGP) and was advertised with a network or aggregate-address command.

e—Path originated from an Exterior Gateway Protocol (EGP).

?—Origin of the path is not clear. Usually, this is a route that is redistributed into BGP from an IGP.

Network

IP prefix and prefix length for a network.

Next Hop

IP address of the next system that is used when a packet is forwarded to the destination network. An entry of 0.0.0.0 indicates that the router has a non-BGP route to this network.

Metric

Value of the interautonomous system metric, otherwise known as the Multi Exit discriminator (MED) metric.

LocPrf

Local preference value. This is used to determine the preferred exit point from the local autonomous system. It is propagated throughout the local autonomous system.

Weight

Path weight. Weight is used in choosing the preferred path to a route. It is not advertised to any neighbor.

Path

Autonomous system path to the destination network. At the end of the path is the origin code for the path.

show bgp session-group

To display information about the Border Gateway Protocol (BGP) configuration for session groups, use the show bgp session-group command in EXECEXEC mode mode.

show bgp session-group group-name {configuration [defaults] [nvgen] | inheritance | users}

Syntax Description

group-name

Name of the session family group to display.

configuration

(Optional) Displays the effective configuration for the session group, including any inherited configuration.

defaults

(Optional) Displays all configuration, including default configuration.

nvgen

(Optional) Displays output in the form of the show running-config command.

If the defaults keyword also is specified, the output is not suitable for cutting and pasting into a configuration session.

inheritance

(Optional) Displays the session groups from which this session group inherits configuration.

users

(Optional) Display the session groups, neighbor groups, and neighbors that inherit configuration from this session group.

Command Default

No default behavior or value

Command Modes

EXECEXEC mode

Command History

Release

Modification

Release 3.7.2

This command was introduced.

Usage Guidelines

To use this command, you must be in a user group associated with a task group that includes appropriate task IDs. If the user group assignment is preventing you from using a command, contact your AAA administrator for assistance.

Use the show bgp session-group command with the group-name configuration argument and keyword to display the effective configuration of a session group, including any configuration inherited from other session groups through application of the use command. The source for each configured command is also displayed.

Use the defaults keyword to display the value of all configuration, including default configuration. Use the nvgen keyword to display configuration in the form of the show running-config command output. Output in this form is suitable for cutting and pasting into a configuration session.

Use the show bgp session-group command with the group-name inheritance argument and keyword to display the session groups from which the specified session group inherits configuration.

Use the show bgp session-group command with the group-name users argument and keyword to display the neighbors, neighbor groups, and session groups that inherit configuration from the specified session group.

Task ID

Task ID

Operations

bgp

read

Examples

For the example shown here, the following configuration is used:


session-group group3
  advertisement-interval 5
  dmzlink-bw
 !
 session-group group1
  use session-group group2
  update-source Loopback0
 !
 session-group group2
  use session-group group3
  ebgp-multihop 2

The following example shows the show bgp session-group command with the configuration keyword:


Router# show bgp session-group group1 configuration  

session-group group1
 advertisement-interval 5[s:group2 s:group3]
 ebgp-multihop 2         [s:group2]
 update-source Loopback0 []
 dmzlink-bandwidth              [s:group2 s:group3]

The source of each command is shown to the right of the command. For example, update-source is configured directly on session group group1. The dmzlink-bandwidth command is inherited from session group group2, which in turn inherits it from session group group3.

The following example shows the show bgp session-group command with the users keyword:


Router# show bgp session-group group2 users

IPv4 Unicast:a:group1

The following example shows the show bgp session-group command with the inheritance keyword.


Router# show bgp session-group group1 inheritance

Session:s:group2 s:group3

The command output shows that the session group group1 directly uses the group2 session group. The group2 session group uses the group3 session group.

This table describes the significant fields shown in the display.

Table 34. show bgp session-group Field Descriptions

Field

Description

[ ]

Configures the command directly on the specified session group.

s:

Indicates the name that follows is a session group.

a:

Indicates the name that follows is an address family group.

n:

Indicates the name that follows is a neighbor group.

[dflt]

Indicates the command is not explicitly configured or inherited, and the default value for the command is used. This field may be shown when the defaults keyword is specified.

<not set>

Indicates that the default is for the command to be disabled. This field may be shown when the defaults keyword is specified.

show bgp sessions

To display brief information about BGP neighbors, use the show bgp sessions command in EXEC mode.

show bgp sessions [not-established] [not-nsr-ready]

Syntax Description

not-established

(Optional) Displays all the neighbors that are not in established state

not-nsr-ready

(Optional) Displays all the neighbors that are not nonstop routing (NSR) ready.

Command Default

No default behavior or values

Command Modes

EXEC

Command History

Release

Modification

Release 3.9.0

This command was introduced.

Usage Guidelines

To use this command, you must be in a user group associated with a task group that includes appropriate task IDs. If the user group assignment is preventing you from using a command, contact your AAA administrator for assistance.

The show bgp sessions command without a keyword provides brief information about all the BGP neighbors configured irrespective of the address family or VRF.

The show bgp sessions command with the not-established keyword shows BGP peers which are yet to establish their peering relationship.

The show bgp session command with the and not-nsr-ready keyword shows BGP peers which are yet to reach the nsr ready state.

Task ID

Task ID

Operations

bgp

read

Examples

The following is sample output from the show bgp sessions command in EXEC mode:


RP/0/RSP0/CPU0:router# show bgp sessions
Thu Jan 15 17:41:45.277 UTC 
 
Neighbor        VRF                   Spk    AS   InQ  OutQ  NBRState     NSRState
2.2.2.2         default                 0     1     0     0  Active       None
10.0.101.1      default                 0     1     0     0  Established  NSR Ready
10.0.101.2      default                 0     1     0     0  Established  NSR Ready
10.0.101.3      default                 0     1     0     0  Established  NSR Ready
10.0.101.4      default                 0     1     0     0  Established  NSR Ready
10.0.101.5      default                 0     1     0     0  Established  NSR Ready
10.0.101.6      default                 0     1     0     0  Established  NSR Ready
10.0.101.7      default                 0     1     0     0  Established  NSR Ready
10.0.101.8      default                 0     1     0     0  Established  NSR Ready
10.0.101.9      default                 0     1     0     0  Established  NSR Ready
10.11.12.2      default                 0   100     0     0  Established  NSR Ready
90.0.0.2        900                     0     2     0     0  Established  NSR Ready
9000::1001      900                     0     2     0     0  Established  NSR Ready
91.0.0.2        901                     0     2     0     0  Established  NSR Ready
9100::1001      901                     0     2     0     0  Established  NSR Ready
92.0.0.2        902                     0     2     0     0  Established  NSR Ready
9200::1001      902                     0     2     0     0  Established  NSR Ready
93.0.0.2        903                     0     2     0     0  Established  NSR Ready
9300::1001      903                     0     2     0     0  Established  NSR Ready
94.0.0.2        904                     0     2     0     0  Established  NSR Ready
9400::1001      904                     0     2     0     0  Established  NSR Ready
95.0.0.2        905                     0     2     0     0  Established  NSR Ready
9500::1001      905                     0     2     0     0  Established  NSR Ready
96.0.0.2        906                     0     2     0     0  Established  NSR Ready
9600::1001      906                     0     2     0     0  Established  NSR Ready
97.0.0.2        907                     0     2     0     0  Established  NSR Ready
9700::1001      907                     0     2     0     0  Established  NSR Ready
98.0.0.2        908                     0     2     0     0  Established  NSR Ready
9800::1001      908                     0     2     0     0  Established  NSR Ready
99.0.0.2        909                     0     2     0     0  Idle         None
9900::1001      909                     0     2     0     0  Idle         None
12.13.14.16     red                     0     2     0     0  Idle         None
20.0.101.1      red                     0     2     0     0  Active       None
1234:5678:9876::1111
                red                     0     3     0     0  Idle         None
2020::1002      red                     0     2     0     0  Established  NSR Ready
1.2.3.4         this-is-a-long-vrf-name
                                        0     5     0     0  Idle         None
1111:2222:3333:4444:5555::6789
                this-is-a-long-vrf-name
                                        0     7     0     0  Idle         None

The following is sample output from the show bgp sessions command with the not-established keyword:


RP/0/RSP0/CPU0:router# show bgp sessions not-established 
Fri Jan 30 11:30:42.720 PST PDT

Neighbor        VRF                   Spk    AS   InQ  OutQ  NBRState     NSRState
10.0.101.5      default                 0   100     0     0  Active       None
2.2.2.2         vrf1_1                  0   302     0     0  Idle         None
2.101.1.2       vrf1_1                  0   302     0     0  Idle         None
2.102.1.2       vrf1_1                  0   302     0     0  Idle         None
2.103.1.2       vrf1_1                  0   302     0     0  Idle         None
4.4.4.2         vrf1_1                  0   304     0     0  Idle         None
2008:2:2:2::2   vrf1_1                  0   302     0     0  Idle         None
11.16.1.2       vrf2_1                  0   302     0     0  Idle         None

The following is sample output from the show bgp sessions command with the not-nsr-ready keyword:


RP/0/RSP0/CPU0:router# show bgp sessions not-nsr-ready 
Fri Jan 30 11:30:52.301 PST PDT

Neighbor        VRF                   Spk    AS   InQ  OutQ  NBRState     NSRState
10.0.101.5      default                 0   100     0     0  Active       None
2.2.2.2         vrf1_1                  0   302     0     0  Idle         None
2.101.1.2       vrf1_1                  0   302     0     0  Idle         None
2.102.1.2       vrf1_1                  0   302     0     0  Idle         None
2.103.1.2       vrf1_1                  0   302     0     0  Idle         None
4.4.4.2         vrf1_1                  0   304     0     0  Idle         None
2008:2:2:2::2   vrf1_1                  0   302     0     0  Idle         None
11.16.1.2       vrf2_1                  0   302     0     0  Idle         None

This table describes the significant fields shown in the display.

Table 35. show bgp sessions Field Descriptions

Field

Description

Neighbor

Displays neighbor IP address.

VRF

Displays information about the VRF.

Spk

Speaker process that is responsible for the neighbor. Always 0.

AS

Autonomous system.

InQ

Number of messages from a neighbor waiting to be processed.

OutQ

Number of messages waiting to be sent to a neighbor.

NBRState

State of the Border Gateway Protocol (BGP) neighbor sessions.

NSRState

State of the Border Gateway Protocol (BGP) nonstop routing (NSR).

show bgp summary

To display the status of all Border Gateway Protocol (BGP) connections, use the show bgp summary command in EXECEXEC mode mode.

show bgp [ipv4 {unicast | multicast | labeled-unicast | all | tunnel | mdt} | ipv6 {unicast | multicast | all | labeled-unicast} | all {unicast | multicast | all | labeled-unicast | mdt | tunnel} | vpnv4 unicast | vrf {vrf-name | all} [ipv4 {unicast | labeled-unicast} | ipv6 unicast] | vpvn6 unicast] summary

Syntax Description

ipv4

(Optional) Specifies IP Version 4 address prefixes.

unicast

(Optional) Specifies unicast address prefixes.

multicast

(Optional) Specifies multicast address prefixes.

labeled-unicast

(Optional) Specifies labeled unicast address prefixes.

all

(Optional) For subaddress families, specifies prefixes for all subaddress families.

tunnel

(Optional) Specifies tunnel address prefixes.

multicast

(Optional) Specifies multicast address prefixes.

ipv6

(Optional) Specifies IP Version 6 address prefixes.

all

(Optional) For address family, specifies prefixes for all address families.

vpnv4 unicast

(Optional) Specifies VPNv4 unicast address families.

vrf

(Optional) Specifies VPN routing and forwarding (VRF) instance.

vrf-name

(Optional) Name of a VRF.

all

(Optional) For VRF, specifies all VRFs.

ipv4 { unicast | labeled-unicast }

(Optional) For VRF, specifies IPv4 unicast or labeled-unicast address families.

ipv6 unicast

(Optional) For VRF, specifies IPv6 unicast address families.

Command Default

If no address family or subaddress family is specified, the default address family and subaddress family specified using the set default-afi and set default-safi commands are used.

Command Modes

EXECEXEC mode

Command History

Release

Modification

Release 3.7.2

This command was introduced.

Release 3.9.0

Asplain format for 4-byte Autonomous system numbers notation was supported. The input parameters and output were modified to display 4-byte autonomous system numbers and extended communities in either asplain or asdot notations.

Usage Guidelines

To use this command, you must be in a user group associated with a task group that includes appropriate task IDs. If the user group assignment is preventing you from using a command, contact your AAA administrator for assistance.


Note


The set default-afi command is used to specify the default address family for the session, and the set default-safi command is used to specify the default subaddress family for the session. See System Management Command Reference for Cisco ASR 9000 Series Routers for detailed information and syntax for the set default-afi and set default-safi commands. If you do not specify a default address family, the default address family is IPv4. If you do not specify a default subaddress family, the default subaddress family is unicast.


Use the show bgp summary command to display a summary of the neighbors for which the specified address family and subaddress family are enabled. If the neighbor does not have the specified address family and subaddress family enabled, it is not included in the output of the show command. If the all keyword is specified for the address family or subaddress family, a summary for each combination of address family and subaddress family is displayed in turn.

The table versions shown in the output (RcvTblVer, bRIB/RIB, SendTblVer, and TblVer) are specific to the specified address family and subaddress family. All other information is global.

The table versions provide an indication of whether BGP is up to date with all work for the specified address family and subaddress family.

  • bRIB/RIB < RecvTblVer—Some received routes have not yet been considered for installation in the global routing table.
  • TblVer < SendTblVer—Some received routes have been installed in the global routing table but have not yet been considered for advertisement to this neighbor.

Task ID

Task ID

Operations

bgp

read

Examples

The following is sample output from the show bgp summary command:

Router#show bgp summary

BGP router identifier 10.0.0.0, local AS number 2
BGP generic scan interval 60 secs
BGP table state: Active
Table ID: 0xe0000000
BGP main routing table version 1
BGP scan interval 60 secs

BGP is operating in STANDALONE mode.

Process         RecvTblVer    bRIB/RIB  LabelVer  ImportVer  SendTblVer
Speaker                  1           0         1          1           0

Neighbor        Spk    AS MsgRcvd MsgSent   TblVer  InQ OutQ Up/Down  St/PfxRcd
10.0.101.0        0     2       0       0        0    0    0 00:00:00 Idle
10.0.101.1        0     2       0       0        0    0    0 00:00:00 Idle

This table describes the significant fields shown in the display.

Table 36. show bgp summary Field Descriptions

Field

Description

BGP router identifier

IP address of the router.

local AS number

Autonomous system number set by the router bgp command.

  • Range for 2-byte Autonomous system numbers (ASNs) is 1 to 65535.
  • Range for 4-byte Autonomous system numbers (ASNs) in asplain format is 1 to 4294967295.
  • Range for 4-byte Autonomous system numbers (ASNs) is asdot format is 1.0 to 65535.65535.

BGP generic scan interval

Interval (in seconds) between scans of the BGP table by a generic scanner.

BGP table state

State of the BGP database.

Table ID

BGP database identifier.

BGP main routing table version

Last version of the BGP database that was injected into the main routing table.

Dampening enabled

Displayed if dampening has been enabled for the routes in this BGP routing table.

BGP scan interval

Interval (in seconds) between scans of the BGP table specified by the address family and subaddress family.

BGP is operating in

Specifies BGP is operating in standalone mode.

Process

BGP process.

RecvTblVer

Last version used in the BGP database for received routes.

bRIB/RIB

Last version of the local BGP database that was injected into the main routing table.

LabelVer

Label version used in the BGP database for label allocation.

ImportVer

Last version of the local BGP database for importing routes.

SendTblVer

Latest version of the local BGP database that is ready to be advertised to neighbors.

Some configured eBGP neighbors do not have any policy

Some external neighbors exist that do not have both an inbound and outbound policy configured for every address family, using the route-policy (BGP) command. In this case, no prefixes are accepted and advertised to those neighbors.

Neighbor

IP address of a neighbor.

Spr

Speaker process that is responsible for the neighbor. Always 0.

AS

Autonomous system.

MsgRcvd

Number of BGP messages received from a neighbor.

MsgSent

Number of BGP messages sent to a neighbor.

TblVer

Last version of the BGP database that was sent to a neighbor.

InQ

Number of messages from a neighbor waiting to be processed.

OutQ

Number of messages waiting to be sent to a neighbor.

Up/Down

Length of time in (hh:mm:ss) that the BGP session has been in Established state, or the time since the session left Established state, if it is not established.

St/PfxRcd

If the BGP session is not established, the current state of the session. If the session is established, the number of prefixes the router has received from the neighbor.

If the number of prefixes received exceeds the maximum allowed (as set by the maximum-prefix command), “(PfxRcd)” appears.

If the connection has been shut down using the shutdown command, “(Admin)” appears.

If the neighbor is external and it does not have an inbound and outbound policy configured for every address family, an exclamation mark (!) is inserted at the end of the state when using the route-policy ( BGP ) command.

If the connection has been shut down due to out of memory (OOM), “(OOM)” appears.

show bgp summary nsr

To display the summary of Border Gateway Protocol (BGP) neighbor state and nonstop routing (NSR) state information, use the show bgp summary nsr command in EXEC mode.

show bgp summary [ipv4 {unicast | multicast | labeled-unicast | all | tunnel | mdt} | ipv6 {unicast | multicast | all | labeled-unicast} | all {unicast | multicast | all | labeled-unicast | mdt | tunnel} | vpnv4 unicast | vrf {vrf-name | all} [ipv4 {unicast | labeled-unicast} | ipv6 unicast] | vpvn6 unicast] nsr [standby]

Syntax Description

ipv4

(Optional) Specifies IP Version 4 address prefixes.

unicast

(Optional) Specifies unicast address prefixes.

multicast

(Optional) Specifies multicast address prefixes.

labeled-unicast

(Optional) Specifies labeled unicast address prefixes.

all

(Optional) For subaddress families, specifies prefixes for all subaddress families.

tunnel

(Optional) Specifies tunnel address prefixes.

mdt

(Optional) Specifies multicast distribution tree (MDT) address prefixes.

multicast

(Optional) Specifies multicast address prefixes.

ipv6

(Optional) Specifies IP Version 6 address prefixes.

all

(Optional) For address family, specifies prefixes for all address families.

vpnv4 unicast

(Optional) Specifies VPNv4 unicast address families.

vrf

(Optional) Specifies VPN routing and forwarding (VRF) instance.

vrf-name

(Optional) Name of a VRF.

all

(Optional) For VRF, specifies all VRFs.

ipv4 { unicast | labeled-unicast }

(Optional) For VRF, specifies IPv4 unicast or labeled-unicast address families.

ipv6 unicast

(Optional) For VRF, specifies IPv6 unicast address families.

vpnv6 unicast

(Optional) Specifies VPNv6 unicast address families.

standby

Displays information about the standby card.

Command Default

If no address family or subaddress family is specified, the default address family and subaddress family specified using the set default-afi and set default-safi commands are used.

Command Modes

EXEC

Command History

Release

Modification

Release 3.9.0

This command was introduced.

Usage Guidelines

To use this command, you must be in a user group associated with a task group that includes appropriate task IDs. If the user group assignment is preventing you from using a command, contact your AAA administrator for assistance.

Task ID

Task ID

Operations

bgp

read

Examples

The following is sample output from the show bgp summary nsr command:


RP/0/RSP0/CPU0:router# show bgp summary nsr 

BGP router identifier 10.1.0.1, local AS number 100
BGP generic scan interval 60 secs
Non-stop routing is enabled
BGP table state: Active
Table ID: 0xe0000000
BGP main routing table version 13037
BGP NSR Initial initsync version 11034 (Reached)
BGP scan interval 60 secs
 
BGP is operating in STANDALONE mode.
 

node0_1_CPU0        Speaker      
 
Entered mode  Standby Ready               : Feb  3 14:22:00
Entered mode  TCP NSR Setup               : Feb  3 14:22:00
Entered mode  TCP NSR Setup Done          : Feb  3 14:22:01
Entered mode  TCP Initial Sync            : Feb  3 14:22:01
Entered mode  TCP Initial Sync Done       : Feb  3 14:22:44
Entered mode  FPBSN processing done       : Feb  3 14:22:44
Entered mode  Update processing done      : Feb  3 14:22:44
Entered mode  BGP Initial Sync            : Feb  3 14:22:44
Entered mode  BGP Initial Sync done       : Feb  3 14:22:49
Entered mode  NSR Ready                   : Feb  3 14:22:49
 
Current BGP NSR state - NSR Ready achieved at: Feb  3 14:22:49
NSR State READY notified to Redcon at: Feb  4 07:44:43
 
Process       RcvTblVer   bRIB/RIB   LabelVer  ImportVer  SendTblVer  StandbyVer
Speaker           13037      13037      13037      13037       13037       13037
 
Neighbor        Spk    AS   TblVer  SyncVer   AckVer NBRState     NSRState
2.2.2.2           0   302    13037    13037    13037 Established  NSR Ready   
10.0.101.5        0   100    13037    13037    13037 Established  NSR Ready 

The following example shows sample output from the show bgp summary nsr command with the standby keyword:


RP/0/RSP0/CPU0:router# show bgp summary nsr standby

BGP router identifier 10.1.0.1, local AS number 100
BGP generic scan interval 60 secs
Non-stop routing is enabled
BGP table state: Active
Table ID: 0xe0000000
BGP main routing table version 13037
BGP NSR Initial initsync version 0 (Not Reached)
BGP scan interval 60 secs
 
BGP is operating in STANDALONE mode.
 

node0_0_CPU0        Speaker      
 
Entered mode  Standby Ready               : Feb  3 14:22:03
Entered mode  TCP Replication             : Feb  3 14:22:03
Entered mode  TCP Init Sync Done          : Feb  3 14:22:47
Entered mode  NSR Ready                   : Feb  3 14:22:52
 
Process       RcvTblVer   bRIB/RIB   LabelVer  ImportVer  SendTblVer  StandbyVer
Speaker           13037          0          0      13037           0           0
 
Neighbor        Spk    AS   TblVer  SyncVer   AckVer NBRState     NSRState
2.2.2.2           0   302    13037        0        1 Established  NSR Ready   
10.0.101.5        0   100    13037        0        1 Established  NSR Ready

This table describes the significant fields shown in the display.

Table 37. show bgp summary nsr Field Descriptions

Field

Description

BGP router identifier

IP address of the router.

BGP generic scan interval

Interval (in seconds) between scans of the BGP table by a generic scanner.

Non-stop routing

State of the Nonstop routing.

BGP table state

State of the BGP database.

Table ID

BGP database identifier.

BGP main routing table version

Last version of the BGP database that was injected into the main routing table.

BGP scan interval

Interval (in seconds) between scans of the BGP table specified by the address family and subaddress family.

BGP is operating in

Specifies BGP is operating in standalone mode.

Entered mode

The successive transition of various states of TCP and BGP, leading to the NSR ready state.

Note

 

This is used for monitoring and debugging purposes.

SyncVer

The version which has synced to standby for this neighbor.

AckVer

The version which the neighbor has acknowledge.

NBRState

State of the BGP neighbor.

NSRState

Neighbor NSR state.

show bgp table

To display the status of all Border Gateway Protocol (BGP) neighbors for a particular Address Family (AF) in the global address table, use the show bgp table command in EXEC mode.

show bgp table [ipv4 {mdt | multicast | mvpn | rt-filter | tunnel | unicast} | ipv6 {multicast | mvpn | unicast} | l2vpn {evpn | vpls | vpws} | standby | vpnv4 unicast | vpnv6 unicast]

Syntax Description

ipv4 mdt

(Optional) Specifies IPv4 multicast distribution tree (MDT) neighbors.

ipv4 multicast

(Optional) Specifies IPv4 multicast neighbors.

ipv4 mvpn

(Optional) Specifies the IPv4 mvpn address family neighbors.

ipv4 rt-filter

(Optional) Specifies the IPv4 RT Constraint address family neighbors.

ipv4 tunnel

(Optional) Specifies IPv4 tunnel neighbors.

ipv6 unicast

(Optional) Specifies IP Version 6 (IPv6) unicast neighbors.

ipv6 multicast

(Optional) Specifies IPv6 multicast neighbors.

ipv6 mvpn

(Optional) Specifies the IPv6 mvpn address family neighbors.

ipv6 unicast

(Optional) Specifies the IPv6 Tunnel address family neighbors.

l2vpn evpn

(Optional) Specifies the L2VPN EVPN address family neighbors.

l2vpn vpls

(Optional) Specifies the L2VPN VPLS address family neighbors.

l2vpn vpws

(Optional) Specifies the L2VPN VPWS address family neighbors.

standby

(Optional) Specifies the IPv4 Unicast address family neighbor on the standby processor.

vpnv4 unicast

(Optional) Specifies VPN Version 4 (VPNv4) unicast address family neighbors.

vpnv6 unicast

(Optional) Specifies VPN Version 6 (VPNv6) unicast address family neighbors.

Command Default

If no address family or subaddress family is specified, the default address family and subaddress family specified using the set default-afi and set default-safi commands are used.

Command Modes

EXEC

Command History

Release

Modification

Release 3.8.0

This command was introduced.

Release 4.3.2, 5.1.0, 5.1.1, 5.1.2 and 5.2.0

The L2VPN Address Family was added.

Usage Guidelines

To use this command, you must be in a user group associated with a task group that includes appropriate task IDs. If the user group assignment is preventing you from using a command, contact your AAA administrator for assistance.

The set default-afi command is used to specify the default address family for the session, and the set default-safi command is used to specify the default subaddress family for the session. See Cisco ASR 9000 Series Aggregation Services Router System Management Command Reference for detailed information and syntax for the set default-afi and set default-safi commands. If you do not specify a default address family, the default address family is IPv4. If you do not specify a default subaddress family, the default subaddress family is unicast.

Use the show bgp table command to display a brief summary of the neighbors for which the specified address family (AFI) and subaddress family (SAFI) are enabled. If the AFI and/or SAFI is not enabled, the command will only display the column headings.

Task ID

Task ID

Operations

bgp

read

Examples

The following is sample output from the bgp table vpnv4 unicast command in EXEC mode:


RP/0/RSP0/CPU0:router# show bgp table vpnv4 unicast 
Thu Jan 15 17:43:31.215 UTC 
Neighbor        VRF                   Spk    AS   TblVer  InQ OutQ  St/PfxRcd
10.0.101.1      default                 0     1      951    0    0         11
10.0.101.2      default                 0     1      951    0    0          5
10.0.101.3      default                 0     1      951    0    0          0
10.0.101.4      default                 0     1      951    0    0          0
10.0.101.5      default                 0     1      951    0    0          0
10.0.101.6      default                 0     1      951    0    0          0
10.0.101.7      default                 0     1      951    0    0          0
10.0.101.8      default                 0     1      951    0    0          0
10.0.101.9      default                 0     1      951    0    0          0
90.0.0.2        900                     0     2      951    0    0          1
91.0.0.2        901                     0     2      951    0    0          1
92.0.0.2        902                     0     2      951    0    0          1
93.0.0.2        903                     0     2      951    0    0          3
94.0.0.2        904                     0     2      951    0    0          3
95.0.0.2        905                     0     2      951    0    0          3
96.0.0.2        906                     0     2      951    0    0          3
97.0.0.2        907                     0     2      951    0    0          3
98.0.0.2        908                     0     2      951    0    0          3
99.0.0.2        909                     0     2        0    0    0 Idle
12.13.14.16     red                     0     2        0    0    0 Idle
20.0.101.1      red                     0     2        0    0    0 Active
1.2.3.4         this-is-a-long-vrf-name
                                        0     5        0    0    0 Idle

This table describes the significant fields shown in the display.

Table 38. show bgp table Field Descriptions

Field

Description

Neighbor

IP address of a neighbor.

VRF

The VRF which each neighbor belongs to; either the default VRF or a specified VRF.

Spk

Speaker process that is responsible for the neighbor. Always 0.

AS

Autonomous system.

TblVer

Last version of the BGP database that was sent to a neighbor.

InQ

Number of messages from a neighbor waiting to be processed.

OutQ

Number of messages waiting to be sent to a neighbor.

St/PfxRcd

If the BGP session is not established, the current state of the session. If the session is established, the number of prefixes the router has received from the neighbor.

If the number of prefixes received exceeds the maximum allowed (as set by the maximum-prefix command), “(PfxRcd)” appears.

If the connection has been shut down using the shutdown command, “(Admin)” appears.

If the neighbor is external and it does not have an inbound and outbound policy configured for every address family, an exclamation mark (!) is inserted at the end of the state when using the route-policy (BGP) command.

If the connection has been shut down due to out of memory (OOM), “(OOM)” appears.

show bgp truncated-communities

To display routes in the Border Gateway Protocol (BGP) routing table for which inbound policy or aggregation has exceeded the maximum number of communities that may be attached, use the show bgp truncated-communities command in EXEC mode mode.

show bgptruncated-communities

Syntax Description

ipv4

(Optional) Specifies IP Version 4 address prefixes.

unicast

(Optional) Specifies unicast address prefixes.

multicast

(Optional) Specifies multicast address prefixes.

labeled-unicast

(Optional) Specifies labeled unicast address prefixes.

all

(Optional) For subaddress families, specifies prefixes for all subaddress families.

tunnel

(Optional) Specifies tunnel address prefixes.

multicast

(Optional) Specifies multicast address prefixes.

ipv6

(Optional) Specifies IP Version 6 address prefixes.

all

(Optional) For address family, specifies prefixes for all address families.

vpnv4 unicast

(Optional) Specifies VPNv4 unicast address families.

rd rd-address

(Optional) Displays routes with a specific route distinguisher.

vrf

(Optional) Specifies VPN routing and forwarding (VRF) instance.

vrf-name

(Optional) Name of a VRF.

all

(Optional) For VRF, specifies all VRFs.

ipv4 { unicast | labeled-unicast }

(Optional) For VRF, specifies IPv4 unicast or labeled-unicast address families.

ipv6 unicast

(Optional) For VRF, specifies IPv6 unicast address families.

Command Default

If no address family or subaddress family is specified, the default address family and subaddress family specified using the set default-afi and set default-safi commands are used.

Command Modes

EXEC mode

Command History

Release

Modification

Release 3.7.2

This command was introduced.

Usage Guidelines

To use this command, you must be in a user group associated with a task group that includes appropriate task IDs. If the user group assignment is preventing you from using a command, contact your AAA administrator for assistance.


Note


The set default-afi command is used to specify the default address family for the session, and the set default-safi command is used to specify the default subaddress family for the session. See the System Management Command Reference for Cisco ASR 9000 Series Routers for detailed information and syntax for the set default-afi and set default-safi commands. If you do not specify a default address family, the default address family is IPv4. If you do not specify a default subaddress family, the default subaddress family is unicast.


BGP contains a separate routing table for each address family and subaddress family combination that has been configured. The address family and subaddress family options specify the routing table to be examined. If the all keyword is specified for the address family or subaddress family, each matching routing table is examined.

Use the show bgp truncated-communities command to display those routes in the specified BGP routing table in which the buffers used to store communities or extended communities have overflowed. An overflow occurs if an attempt is made to associate more communities or extended communities with the route than fits in a BGP update message. This can happen due to modification of communities or extended communities during aggregration or when inbound policy is applied.

Task ID

Task ID

Operations

bgp

read

Examples

The following is sample output from the show bgp truncated-communities command:


Router# show bgp truncated-communities

BGP router identifier 172.20.1.1, local AS number 1820
BGP main routing table version 3042
BGP scan interval 60 secs
Status codes: s suppressed, d damped, h history, * valid, > best
              i - internal, S stale
Origin codes: i - IGP, e - EGP, ? - incomplete
   Network            Next Hop            Metric LocPrf Weight Path
*  10.13.0.0/16       192.168.40.24                          0 1878 704 701 200 ?
*> 10.16.0.0/16       192.168.40.24                          0 1878 704 701 i

This table describes the significant fields shown in the display.

Table 39. show bgp truncated-communities Field Descriptions

Field

Description

BGP router identifier

BGP Identifier for the local system.

local AS number

Autonomous system number for the local system.

BGP main routing table version

Last version of the BGP database that was installed into the main routing table.

Dampening enabled

Displayed if dampening is enabled for the routes in this BGP routing table.

BGP scan interval

Interval (in seconds) between scans of the BGP table specified by the address family and subaddress family.

Status codes

Status of the table entry. The status is displayed as a three-character field at the beginning of each line in the table. The first character may be (in order of precedence):

S—Path is stale, indicating that a graceful restart is in progress with the peer from which the route was learned.

s—Path is more specific than a locally sourced aggregate route and has been suppressed.

*—Path is valid.

The second character may be (in order of precedence):

>—Path is the best path to use for that network.

d—Path is dampened.

h—Path is a history entry, representing a route that is currently withdrawn, but that is being maintained to preserve dampening information. Such routes should never be marked as valid.

The third character may be:

i—Path was learned by an internal BGP (iBGP) session.

Origin codes

Origin of the path. The origin code is displayed at the end of each line in the table. It can be one of the following values:

i—Path originated from an Interior Gateway Protocol (IGP) and was advertised with a network or aggregate-address command.

e—Path originated from an Exterior Gateway Protocol (EGP).

?—Origin of the path is not clear. Usually, this is a route that is redistributed into BGP from an IGP.

Network

IP prefix and prefix length for a network.

Next Hop

IP address of the next system that is used when a packet is forwarded to the destination network. An entry of 0.0.0.0 indicates that the router has a non-BGP route to this network.

Metric

Value of the interautonomous system metric, otherwise known as the Multi Exit Discriminator (MED) metric.

LocPrf

Local preference value. This is used to determine the preferred exit point from the local autonomous system. It is propagated throughout the local autonomous system.

Weight

Path weight. Weight is used in choosing the preferred path to a route. It is not advertised to any neighbor.

Path

Autonomous system path to the destination network. At the end of the path is the origin code for the path.

show bgp update-group

To display Border Gateway Protocol (BGP) information for update groups, use the show bgp update-group command in EXEC mode.

show bgp [ipv4 {unicast | multicast | labeled-unicast | all | tunnel | mdt} | ipv6 {unicast | multicast | all | labeled-unicast} | all {unicast | multicast | all | labeled-unicast | mdt | tunnel} | vpnv4 unicast | vrf {vrf-name | all} [ipv4 {unicast | labeled-unicast} | ipv6 unicast] | vpnv6 unicast] update-group [neighbor ip-address | process-id. index [summary | performance-statistics]]

Syntax Description

ipv4

(Optional) Specifies IP Version 4 update groups.

unicast

(Optional) Specifies unicast update groups.

multicast

(Optional) Specifies multicast update groups.

labeled-unicast

(Optional) Specifies labeled unicast address prefixes.

all

(Optional) Displays both unicast and multicast update groups.

tunnel

(Optional) Specifies tunnel address prefixes.

ipv6

(Optional) Specifies IP Version 6 update groups.

all

(Optional) Displays both IP Version 4 and IP Version 6 update groups.

vpnv4 unicast

(Optional) Specifies VPNv4 unicast address families.

rd rd-address

(Optional) Displays routes with a specific route distinguisher.

vrf

(Optional) Specifies VPN routing and forwarding (VRF) instance.

vrf-name

(Optional) Name of a VRF.

all

(Optional) For VRF, specifies all VRFs.

ipv4 { unicast | labeled-unicast }

(Optional) For VRF, specifies IPv4 unicast or labeled-unicast address families.

ipv6 unicast

(Optional) For VRF, specifies IPv6 unicast address families.

neighbor ip-address

(Optional) Specifies information on an update group for a specific neighbor.

process-id.index

(Optional) Update group index. Process ID range is 0 to 254. Index range is 0 to 4294967295.

The process id.index argument is specified as follows: process ID (dot) index. In standalone mode, the process ID is always 0.

summary

(Optional) Specifies summary of update group members.

performance-statistics

(Optional) Specifies performance information about the updates generated for the update group.

Command Default

If no address family or subaddress family is specified, the default address family and subaddress family specified using the set default-afi and set default-safi commands are used.

Command Modes

EXEC EXEC

Command History

Release

Modification

Release 3.7.2

This command was introduced.

Release 3.9.0

Asplain format for 4-byte Autonomous system numbers notation was supported. The input parameters and output were modified to display 4-byte autonomous system numbers and extended communities in either asplain or asdot notations.

Release 7.1.1

The command output was modified to include the performance statistics for BGP peer that was not advertised..

Usage Guidelines


Note


The set default-afi command is used to specify the default address family for the session, and the set default-safi command is used to specify the default subaddress family for the session. See the System Management Command Reference for Cisco ASR 9000 Series Routers for detailed information and syntax for the set default-afi and set default-safi commands. If you do not specify a default address family, the default address family is IPv4. If you do not specify a default subaddress family, the default subaddress family is unicast.


Every BGP neighbor is automatically assigned to an update group for each address family that is enabled on the neighbor. Neighbors that have similar outbound policy, such that they are sent the same updates, are placed in the same update group.

Use the show bgp update-group command to display the update groups and a list of the neighbors that belong to the update group.

Use the show bgp update-group neighbor command to display details about the update group to which a neighbor belongs for the specified address family.

Use the summary keyword to display a summary of the neighbors belonging to the specified update group. The display format is the same as for the show bgp summary command.

Use the performance-statistics keyword to display information about the number of prefixes processed and the time taken to generate updates for the specified update group.


Note


Update group indexes are not necessarily persistent over a process restart. If a BGP process restarts, the index of the update group to which a particular neighbor is assigned may be different, though the set of neighbors belonging to the update group is the same.


Task ID

Task ID

Operations

bgp

read

Examples

The following is sample output from the show bgp update-group command:


RP/0/RSP0/CPU0:router# show bgp update-group

Update group for IPv4 Unicast, index 0.1:
  Attributes:
    Internal
    Common admin
    Send communities
    Send extended communities
    Minimum advertisement interval: 300
  Update group desynchronized: 0
  Sub-groups merged: 0
  Messages formatted: 0, replicated: 0
  Neighbors not in any sub-group:
    10.0.101.1 

The following is sample output from the show bgp update-group performance-statistics command:


RP/0/RSP0/CPU0:router# show bgp update-group performace-statistics


Update group for IPv4 Unicast, index 0.1:
..
Update timer last processed: Sep 23 00:10:15.350
  Not-Advertised Stats:
    Non-Client to Non-Client     : 105        Sep 23 00:10:15.350
    Path Not Available	    : 132        Sep 23 00:10:15.350

The following is sample output from the show bgp update-group performance-statistics command:


Router# show bgp update-group performace-statistics


Update group for IPv4 Unicast, index 0.1:
..
Update timer last processed: Sep 23 00:10:15.350
  Not-Advertised Stats:
    Non-Client to Non-Client     : 105        Sep 23 00:10:15.350
    Path Not Available	    : 132        Sep 23 00:10:15.350

This table describes the significant fields shown in the display.

Table 40. show bgp update-group Field Descriptions

Field

Description

Update group for

Address family to which updates in this update group apply.

index

Update group index.

Attributes

Attributes common to all members of the update group.

Unsuppress map

Unsuppress route map used to selectively unsuppress more specific routes of locally generated aggregates for members of this update group.

Outbound policy

Route policy applied to outbound updates generated for members of this update group.

Internal

Members of the update group are internal peers.

ORF Receive enabled

Members of this update group are capable of receiving an outbound route filter.

Route Reflector Client

Local system is acting as a route reflector for members of this update group.

Remove private AS numbers

Members of this update group have private AS numbers stripped from outbound updates.

Next-hop-self enabled

Next- Next hop for members of the update group is set to the local router.

Directly connected IPv6 EBGP

Members of this update group are directly connected external BGP IPv6-based peers.

Configured Local AS

Local autonomous system (AS) used for members of this update group.

Common admin

Peers in this update group are under common administration (internal or confederation peers).

Send communities

Communities are sent to neighbors in this update group.

Send extended communities

Extended communities is sent to neighbors in this update group.

Minimum advertisement interval

Minimum advertisement interval for members of this update group.

replicated

Number of update messages replicated for this update group.

Messages formatted

Number of update messages generated for this update group.

Neighbors in this update group

List of neighbors that use this update group for the given address family.

Update group desynchronized

Number of times an update group has been split to accommodate the slower peer. This option is disabled.

Sub-groups merged

Number of times an update group has been split and merged.

Neighbors not in any sub-group

BGP neighbor that does not belong to any subgroup.

The following is sample output from the show bgp update-group command with the ipv4 , unicast , and summary keywords and the process id.index argument:


RP/0/RSP0/CPU0:router# show bgp ipv4 unicast update-group 0.1 summary 

BGP router identifier 10.140.140.1, local AS number 1.1
BGP generic scan interval 60 secs
BGP table state: Active
Table ID: 0xe0000000
BGP main routing table version 1
BGP scan interval 60 secs

BGP is operating in STANDALONE mode.

Process         RecvTblVer    bRIB/RIB  LabelVer  ImportVer  SendTblVer
Speaker                  1           0         1          1           0

Neighbor        Spr    AS MsgRcvd MsgSent   TblVer  InQ OutQ Up/Down  St/PfxRcd

172.25.11.8       0     1       0       0        0    0    0 00:00:00 Idle

This table describes the significant fields shown in the display.

Table 41. show bgp ipv4 unicast update-group Field Descriptions

Field

Description

BGP router identifier

IP address of the router.

local AS number

Autonomous system number set by the router bgp command.

  • Range for 2-byte Autonomous system numbers (ASNs) is 1 to 65535.
  • Range for 4-byte Autonomous system numbers (ASNs) in asplain format is 1 to 4294967295.
  • Range for 4-byte Autonomous system numbers (ASNs) is asdot format is 1.0 to 65535.65535.

BGP generic scan interval

Interval (in seconds) between scans of the BGP table by a generic scanner.

BGP table state

State of the BGP database.

Table ID

BGP database identifier.

BGP main routing table version

Last version of the BGP database that was injected into the main routing table.

Dampening enabled

Displayed if dampening has been enabled for the routes in this BGP routing table.

BGP scan interval

Interval (in seconds) between scans of the BGP table specified by the address family and subaddress family.

BGP is operating in

BGP is operating in standalone mode.

Process

BGP process.

RecvTblVer

Last version used in the BGP database for received routes.

bRIB/RIB

Last version of the local BGP database that was injected into the main routing table.

LabelVer

Label version used in the BGP database for label allocation.

ImportVer

Last version of the local BGP database for importing routes.

SendTblVer

Latest version of the local BGP database that is ready to be advertised to neighbors.

Some configured eBGP neighbors do not have any policy

Some external neighbors that exist do not have both an inbound and outbound policy configured for every address family, using the route-policy (BGP) command. In this case, no prefixes are accepted or advertised to those neighbors.

Neighbor

IP address of a neighbor.

Spr

Speaker process that is responsible for the neighbor. Always 0.

AS

Autonomous system.

MsgRcvd

Number of BGP messages received from a neighbor.

MsgSent

Number of BGP messages sent to a neighbor.

TblVer

Last version of the BGP database that was sent to a neighbor.

InQ

Number of messages from a neighbor waiting to be processed.

OutQ

Number of messages waiting to be sent to a neighbor.

Up/Down

Length of time (in hh:mm:s) that the BGP session has been in Established state, or the time since the session left Established state, if it is not established.

St/PfxRcd

If the BGP session is not established, the current state of the session. If the session is established, the number of prefixes the router has received from the neighbor.

If the number of prefixes received exceeds the maximum allowed (as set by the maximum-prefix command), “(PfxRcd)” appears.

If the connection has been shut down using the shutdown command, “(Admin)” appears.

If the neighbor is external and it does not have an inbound and outbound policy configured for every address family, an exclamation mark (!) is inserted at the end of the state when using the route-policy ( BGP ) command.

Examples

This is sample output from the show bgp update-group performance-statistics command that shows the total numbers of prefixes not advertising in any AFI or SAFI, including repeating counts on 1 or more prefixes:
Router# show bgp update-group performance-statistics

Update group for IPv4 Unicast, index 0.1:
..
Update timer last processed: Sep 23 00:10:15.350
  Not-Advertised Stats:
    Non-Client to Non-Client       : 105        Sep 23 00:10:15.350
    Path Not Available	      : 132        Sep 23 00:10:15.350

show bgp vrf

To display Border Gateway Protocol (BGP) prefix information for VPN routing and forwarding (VRF) instances, use the show bgp vrf command in EXEC mode.

show bgp vrf { all | vrf-name } { ipv4 { unicast [ ipv4-address/length [ detail ] ] | labeled-unicast } | ipv6 { unicast } | imported-routes { neighbor | standby | vrf vrf-name }}

Syntax Description

vrf-name

Displays imported routes for a specific VRF.

all

Displays imported routes for all VRFs.

ipv4 { unicast | labeled-unicast }

(Optional) Specifies IP Version 4 unicast or labeled-unicast routes.

ipv4-address/length [ detail ]

(Optional) Displays detailed output for the specified IPv4-address.

ipv6 unicast

(Optional) Specifies IP Version 6 unicast routes.

vrf source-vrf-name

(Optional) Displays routes imported from the specified source VRF.

neighbor neighbor-address

(Optional) Displays preview advertisements for a specified neighbor.

Command Default

No default behavior or values

Command Modes

EXEC mode

Command History

Release

Modification

Release 3.7.2

This command was introduced.

Release 3.9.0

Asplain format for 4-byte Autonomous system numbers notation was supported. The input parameters and output were modified to display 4-byte autonomous system numbers and extended communities in either asplain or asdot notations.

Usage Guidelines

To use this command, you must be in a user group associated with a task group that includes appropriate task IDs. If the user group assignment is preventing you from using a command, contact your AAA administrator for assistance.

Use the show bgp vrf imported-routes command to display all paths imported into a specified VRF from the default VRF. Use the neighbor neighbor-address keyword and argument to display all imported paths and which paths were learned from the specified neighbor. Use the vrf source-vrf-name keyword and argument to display all imported routes that belong to the specified source VRF. The neighbor neighbor-address and vrf source-vrf-name cannot coexist.

Task ID

Task ID

Operations

bgp

read

Examples

The following is sample output from the show bgp vrf imported-routes command:


RP/0/RSP0/CPU0:router# show bgp vrf vrf-1 ipv6 unicast imported-routes

BGP VRF one, state: Active BGP 
BGP Route Distinguisher: 100:222 
VRF ID: 0x60000001 
BGP router identifier 10.2.0.1, local AS number 100 
BGP table state: Active 
Table ID: 0xe0800001 
BGP main routing table version 41534

Status codes: s suppressed, d damped, h history, * valid, > best
              i - internal, S stale
Origin codes: i - IGP, e - EGP, ? - incomplete
   Network            Neighbor        Route Distinguisher    Source VRF
*>i1234:1052::/32     10.1.0.1        100:111                default            
*>i2008:1:1:1::/112   10.1.0.1        100:111                default            
*>i2008:111:1:1::1/128
                      10.1.0.1        100:111                default

Processed 3 prefixes, 3 paths

This table describes the significant fields shown in the display output for show bgp vrf command.

Table 42. show bgp vrf Field Descriptions

Field

Description

BGP VRF

VRF name.

state

State of the VRF.

BGP Route Distinguisher:

Unique identifier for the BGP routing instance.

VRF Id

VRF identifier.

BGP router identifier

IP address of the router.

local AS number

Autonomous system number set by the router bgp command.

  • Range for 2-byte Autonomous system numbers (ASNs) is 1 to 65535.
  • Range for 4-byte Autonomous system numbers (ASNs) in asplain format is 1 to 4294967295.
  • Range for 4-byte Autonomous system numbers (ASNs) is asdot format is 1.0 to 65535.65535.

BGP table state

State of the BGP database.

Table ID

Table identifier.

BGP main routing table version

Last version of the BGP database that was injected into the main routing table.

Network

Network address.

Neighbor

IP address of a neighbor.

Route Distinguisher

Unique identifier for the routing instance.

Source VRF

Source VRF for the imported route.

show lpts punt excessive-flow-trap bgp

To display the details of bad actor identified for bgp protocol, use the show lpts punt excessive-flow-trap bgp command in the Global Configuration mode.

show lpts punt excessive-flow-trap bgp

Command Default

None

Command Modes

Global Configuration mode

Command History

Release Modification

Release 6.0.1

This command was introduced.

Usage Guidelines

You must be in a user group associated with a task group that includes the proper task IDs. The command reference guides include the task IDs required for each command. If you suspect user group assignment is preventing you from using a command, contact your AAA administrator for assistance.

Task ID

Task ID

Operations

lpts

read

basic-services

read-write

Examples

This example shows to the details of bad actor identified for bgp protocol:

RP/0/RSP0/CPU0:router# show lpts punt excessive-flow-trap bgp
Parent Interface: GigabitEthernet0/2/0/19           Src MAC Addr: 0000.6505.0102                  
         Intf Handle: 0x08000580                            Location: 0/2/CPU0
            Protocol: BGP                                         Punt Reason: BGP-mc-known
        Penalty Rate: 0 pps (all packets dropped)    Penalty Timeout: 15 mins                         
      Time Remaining: 10 mins 3 secs

show protocols (BGP)

To display information about the Border Gateway Protocol (BGP) instances running on the router, use the show protocols command in EXEC mode and specify either the bgp or all keyword.

show protocols [ipv4 | ipv6 | afi-all] [all | protocol]

Syntax Description

ipv4

(Optional) Specifies the IP Version 4 address family.

ipv6

(Optional) Specifies the IP Version 6 address family.

afi-all

(Optional) Specifies all address families.

all

(Optional) Specifies all protocols for a given address family.

protocol

(Optional) Specifies a routing protocol.

For the IPv4 address family, the options are bgp , isis , rip , eigrp , and ospf .

For the IPv6 address family, the options are bgp , eigrp , isis , and ospfv3 .

Command Default

Default is IPv4.

Command Modes

EXEC

Command History

Release

Modification

Release 3.7.2

This command was introduced.

Release 3.9.0

Asplain format for 4-byte Autonomous system numbers notation was supported. The input parameters and output were modified to display 4-byte autonomous system numbers and extended communities in either asplain or asdot notations.

Usage Guidelines

To use this command, you must be in a user group associated with a task group that includes appropriate task IDs. If the user group assignment is preventing you from using a command, contact your AAA administrator for assistance.

Use the show protocols command to get information about the protocols running on the router and to quickly determine which protocols are active. The command is designed to summarize the important characteristics of the running protocol, and command output varies depending on the specific protocol selected. For BGP, the command output lists the protocol ID, peers with elapsed time since last reset, and miscellaneous information, such as external and internal local distances and sourced routes.

Task ID

Task ID

Operations

bgp

read

rib

read

Examples

The following example shows the display for the show protocols command using the bgp keyword:


RP/0/RSP0/CPU0:router# show protocols bgp

Routing Protocol "BGP 40"

Address Family IPv4 Unicast:
  Distance: external 20 internal 200 local 200
  Sourced Networks:
    10.100.0.0/16 backdoor
    10.100.1.0/24
    10.100.2.0/24
  Routing Information Sources:
    Neighbor          State/Last update received
    10.5.0.2          Idle
    10.9.0.3          Idle

This table describes the significant fields shown in the display.

Table 43. show protocols (BGP) Field Descriptions

Field

Description

Routing Protocol:

Identifies BGP as the running protocol and displays the BGP AS number.

  • Range for 2-byte Autonomous system numbers (ASNs) is 1 to 65535.
  • Range for 4-byte Autonomous system numbers (ASNs) in asplain format is 1 to 4294967295.
  • Range for 4-byte Autonomous system numbers (ASNs) is asdot format is 1.0 to 65535.65535.

Address Family

Specifies the address family. This can be IPv4 Unicast, IPv4 Multicast, or IPv6 Unicast.

Distance: external

Specifies the distance BGP sets when installing eBGP routes into the RIB. eBGP routes are routes received from eBGP peers. The RIB uses the distance as a tiebreaker when several protocols install a route for the same prefix.

Distance: internal

Specifies the distance BGP sets for routes received from iBGP peers.

Distance: local

Specifies the distance BGP sets for locally generated aggregates and backdoor routes.

Sourced Networks

List of locally sourced networks. These are networks sourced using the network command.

Routing information Sources

List of configured BGP neighbors.

Neighbor

Address of a BGP neighbor.

State/Last update received

State of each neighbor and the time since the last update was received from the neighbor if it is established.

show running-config lpts punt excessive-flow-trap

To display the running configuration for the Excessive Punt Flow Trap feature, use the show running-config lpts punt excessive-flow-trap command in the EXEC mode.

show running-configlpts punt excessive-flow-trap

Command Default

None

Command Modes

EXEC mode

Command History

Release Modification

Release 5.2.2

This command was introduced.

Usage Guidelines

No specific guidelines impact the use of this command.

Task ID

Task ID Operation

lpts

read

basic-services

read, write

Examples

The show running-config output for the above show lpts punt excessive-flow-trap command is:

RP/0/RSP0/CPU0:router# show running-config lpts punt excessive-flow-trap
lpts punt excessive-flow-trap
penalty-rate arp 10
penalty-timeout arp 20
non-subscriber-interfaces
!

This table describes the significant fields shown in the display.

Table 44. show lpts punt excessive-flow-trap Field Descriptions

Field

Description

penalty-rate

The penalty policing rate for the ARP protocol. For ARP, the value is 10.

penalty-timeout

The penalty timeout value for the ARP protocol. For ARP, the value is 20.

show svd role

To display selective VRF download (SVD) role information, use the show svd role command in EXEC mode.

show svd role

Syntax Description

This command has no keywords or arguments.

Command Default

None.

Command Modes

EXEC

Command History

Release Modification

Release 4.2.0

This command was introduced.

Usage Guidelines

To use this command, you must be in a user group associated with a task group that includes appropriate task IDs. If the user group assignment is preventing you from using a command, contact your AAA administrator for assistance.

The show svd role command output displays name of the line card and role for each address-family in a table.

Task ID

Task ID Operation

ip-services

read

Examples

This example displays the different nodes in a line card and corresponding IPv4, and IPv6 SVD role information:


RP/0/RSP0/CPU0:router#show svd role 
Thu Mar 10 10:45:17.886 PST
Node Name       IPv4 Role            IPv6 Role      
-----------------------------------------------------
0/1/CPU0        Core Facing          Not Interested   
0/2/CPU0        Core Facing          Core Facing    
0/4/CPU0        Standard             Standard       
0/5/CPU0        Standard             Standard

show svd state

To display selective VRF download (SVD) state information, use the show svd state command in EXEC mode.

show svd state

Syntax Description

This command has no keywords or arguments.

Command Default

None.

Command Modes

EXEC

Command History

Release Modification

Release 4.2.0

This command was introduced.

Usage Guidelines

To use this command, you must be in a user group associated with a task group that includes appropriate task IDs. If the user group assignment is preventing you from using a command, contact your AAA administrator for assistance.

Task ID

Task ID Operation

ip-services

read

Examples

This example shows the SVD configuration state and the SVD operational state in a line card:


RP/0/RSP0/CPU0:router#show svd state 
Thu Mar 10 10:45:32.184 PST
Selective VRF Download (SVD) Feature State:
SVD Configuration State        Enabled
SVD Operational State          Enabled

show tcp pcb

To display TCP connection information, use the show tcp pcb command in the EXEC mode.

show tcp pcb pcb-value

Syntax Description

pcb-value

Specifies PCB hexadecimal address. The valid range is from 0x0 to 0xFFFFFFFF.

Command Default

No default behavior or values

Command Modes

EXEC mode

Command History

Release Modification

Release 5.3.2

This command was introduced.

Usage Guidelines

No specific guidelines impact the use of this command.

Task ID

Task ID Operation

bgp

read

Examples

The following is a sample output from the show tcp pcb command:



RP/0/0/CPU0:ios#show tcp pcb 0xecb0c9f8

Connection state is ESTAB, I/O status: 0, socket status: 0
Established at Sun Dec 7 11:49:39 2014

PCB 0xecb0c9f8, SO 0xecb01b68, TCPCB 0xecb01d78, vrfid 0x60000000, 
Pak Prio: Medium, TOS: 192, TTL: 255, Hash index: 1322
Local host: 10.0.0.1, Local port: 12404 (Local App PID: 19840)
Foreign host: 10.0.0.2, Foreign port: 179

Current send queue size in bytes: 0 (max 24576)
Current receive queue size in bytes: 0 (max 32768) mis-ordered: 0 bytes
Current receive queue size in packets: 0 (max 0)

Timer Starts Wakeups Next(msec)
Retrans 17 2 0
SendWnd 0 0 0
TimeWait 0 0 0
AckHold 13 5 0
KeepAlive 1 0 0
PmtuAger 0 0 0
GiveUp 0 0 0
Throttle 0 0 0

iss: 1728179225 snduna: 1728179536 sndnxt: 1728179536
sndmax: 1728179536 sndwnd: 32517 sndcwnd: 1000 
irs: 2055835995 rcvnxt: 2055836306 rcvwnd: 32536 rcvadv: 2055868842

SRTT: 206 ms, RTTO: 300 ms, RTV: 59 ms, KRTT: 0 ms
minRTT: 10 ms, maxRTT: 230 ms

ACK hold time: 200 ms, Keepalive time: 0 sec, SYN waittime: 30 sec
Giveup time: 0 ms, Retransmission retries: 0, Retransmit forever: FALSE
Connect retries remaining: 30, connect retry interval: 30 secs

State flags: none
Feature flags: Win Scale, Nagle
Request flags: Win Scale

Datagrams (in bytes): MSS 500, peer MSS 1460, min MSS 500, max MSS 1460 

Window scales: rcv 0, snd 0, request rcv 0, request snd 0
Timestamp option: recent 0, recent age 0, last ACK sent 0
Sack blocks {start, end}: none
Sack holes {start, end, dups, rxmit}: none

Socket options: SO_REUSEADDR, SO_REUSEPORT, SO_NBIO
Socket states: SS_ISCONNECTED, SS_PRIV
Socket receive buffer states: SB_DEL_WAKEUP
Socket send buffer states: SB_DEL_WAKEUP
Socket receive buffer: Low/High watermark 1/32768 
Socket send buffer : Low/High watermark 2048/24576, Notify threshold 0 

PDU information:
#PDU's in buffer: 0
FIB Lookup Cache: IFH: 0x200 PD ctx: size: 0 data: 
Num Labels: 0 Label Stack:

shutdown (BGP)

To disable a neighbor without removing its configuration, use the shutdown command in an appropriate configuration mode. To re-enable the neighbor and reestablish a Border Gateway Protocol (BGP) session, use the no form of this command.

shutdown [inheritance-disable]

no shutdown [inheritance-disable]

Syntax Description

inheritance-disable

(Optional) Overrides the value of a shutdown command inherited from a neighbor group or session group.

Command Default

Neighbors are not shutdown.

Command Modes

Neighbor configuration

VRF neighbor configuration

Neighbor group configuration

Session group configuration

Command History

Release

Modification

Release 3.7.2

This command was introduced.

Release 3.9.0

The disable keyword was replaced with the inheritance-disable keyword.

Usage Guidelines

To use this command, you must be in a user group associated with a task group that includes appropriate task IDs. If the user group assignment is preventing you from using a command, contact your AAA administrator for assistance.

Use the shutdown command to terminate any active session for the specified neighbor and remove all associated routing information. Use of the shutdown command with a neighbor group or session group may suddenly terminate a large number of BGP neighbor sessions because all neighbors using the neighbor group or session group may be affected.

Use the show bgp summary command to display a summary of BGP neighbors. Neighbors that are idle due to the shutdown command are displayed with the “Idle (Admin)” state.

If this command is configured for a neighbor group or session group, all neighbors using the group inherit the configuration. Values of commands configured specifically for a neighbor override inherited values.

Task ID

Task ID

Operations

bgp

read, write

Examples

The following example shows that any active session for neighbor 192.168.40.24 is disabled:


Router(config)# router bgp 1
Router(config-bgp)# neighbor 192.168.40.24
Router(config-bgp-nbr)# shutdown
Router(config-bgp-nbr)# exit

In the following example, the session remains active for neighbor 192.168.40.24 because the inherited shutdown command has been overridden:


Router(config)# router bgp 1
Router(config-bgp)# session-group group1
Router(config-bgp-sngrp)# shutdown
Router(config-bgp-sngrp)# exit
Router(config-bgp)# neighbor 192.168.40.24
Router(config-bgp-nbr)# remote-as 1
Router(config-bgp-nbr)# use session-group group1
Router(config-bgp-nbr)# shutdown inheritance-disable
Router(config-bgp-nbr)# exit

shutdown (rpki-server)

To shutdown RPKI cache-server, use the shutdown command in rpki-server configuration mode. To set that the RPKI cache be active, use the no form of this command.

shutdown

no shutdown

This command has no keywords or arguments.

Command Default

RPKI cache is active.

Command Modes

RPKI server configuration

Command History

Release Modification

Release 4.2.1

This command was introduced.

Usage Guidelines

To use this command, you must be in a user group associated with a task group that includes appropriate task IDs. If the user group assignment is preventing you from using a command, contact your AAA administrator for assistance.

Task ID

Task ID Operation

bgp

read, write

Examples

This command shows how to configure no shutdown of the RPKi cache configuration after other RPKI cache parameters are configured:


Router#configure
Router(config)#router bgp 100
Router(config-bgp)#rpki server 172.168.35.40 
Router(config-bgp-rpki-server)# transport ssh port 22
Router(config-bgp-rpki-server)#username rpki-user
Router(config-bgp-rpki-server)#password rpki-ssh-pass
Router(config-bgp-rpki-server)#preference 1
Router(config-bgp-rpki-server)#purge-time 30
Router(config-bgp-rpki-server)#refresh-time 30
Router(config-bgp-rpki-server)#response-time 30
Router(config-bgp-rpki-server)#no shutdown

signalling disable

To disable BGP or LDP signaling protocol to neighbors, use the signalling disable command in neighbor address family (l2vpn vpls-vpws) configuration mode. To restore the system to its default condition, use the no form of this command.

signalling {bgp | ldp} disable

Syntax Description

bgp

Selects BGP signaling protocol to disable.

ldp

Selects LDP signaling protocol to disable.

Command Default

Both BGP and LDP signaling protocols are enabled.

Command Modes

Neighbor address family configuration

Command History

Release Modification
Release 3.9.1

This command was introduced.

Usage Guidelines

To use this command, you must be in a user group associated with a task group that includes appropriate task IDs. If the user group assignment is preventing you from using a command, contact your AAA administrator for assistance.

Task ID

Task ID Operation
bgp

read, write

Examples

The following example shows how to disable BGP signaling protocol for neighbor 10.2.3.4:


RP/0/RSP0/CPU0:router(config)#router bgp 100
RP/0/RSP0/CPU0:router(config-bgp)#neighbor 10.2.3.4
RP/0/RSP0/CPU0:router(config-bgp-nbr)#address-family l2vpn vpls-vpws 
RP/0/RSP0/CPU0:router(config-bgp-nbr-af)#signalling bgp disable

site-of-origin (BGP)

To attach a site-of-origin extended community attribute to each route received from the specified peer, use the site-of-origin command in VRF neighbor address family configuration mode. To restore the system to its default condition, use the no form of this command.

site-of-origin [as-number:nn | ip-address:nn]

Syntax Description

as-number:nn

  • as-number— Autonomous system (AS) number.
    • Range for 2-byte Autonomous system number is 1 to 65535.

    • Range for 4-byte Autonomous system number in asplain format is 1 to 4294967295.

    • Range for 4-byte Autonomous system number is asdot format is 1.0 to 65535.6553.

  • nn —32-bit number

ip-address:nn

IP address.

  • ip-address —32-bit IP address
  • nn —16-bit number

Command Default

No default behavior or values

Command Modes

VRF neighbor address family configuration

Command History

Release

Modification

Release 3.7.2

This command was introduced.

Release 3.9.0

Asplain format for 4-byte Autonomous system numbers notation was supported.

Usage Guidelines

To use this command, you must be in a user group associated with a task group that includes appropriate task IDs. If the user group assignment is preventing you from using a command, contact your AAA administrator for assistance.

When routes are advertised to the peer, routes whose extended communities list contain the site of origin (SoO) are filtered out and not advertised to the peer. Site-of-origin uniquely identifies the site from which the provide edge (PE) router learned routes, thus filtering based on the extended community helps prevent transient routing loops from occurring in complex and mixed network topologies.

Task ID

Task ID

Operations

bgp

read, write

Examples

The following example shows how to configure SoO filtering:


RP/0/RSP0/CPU0:router(config)# router bgp 6 
RP/0/RSP0/CPU0:router(config-bgp)# vrf vrf_A
RP/0/RSP0/CPU0:router(config-bgp-vrf)# neighbor 192.168.70.24 
RP/0/RSP0/CPU0:router(config-bgp-vrf-nbr)# remote-as 10
RP/0/RSP0/CPU0:router(config-bgp-vrf-nbr)# address-family ipv4 unicast
RP/0/RSP0/CPU0:router(config-bgp-vrf-nbr-af)# site-of-origin 10.0.01:20 

slow peer (BGP router global configuration)

To enable BGP global slow peer configuration, use the slow-peer command in BGP router configuration mode. Enabling global slow peer configuration enables slow peer configuration for all (default VRF and non-default VRF) BGP neighbor address-families.

The default global slow peer mode is detection-only.

Use the no form of the command to return to the default slow peer configuration.

slow-peer {dynamic [threshold sec] | detection-disable}

Syntax Description

dynamic

Enables dynamic slow peer configuration on all (default VRF and non-default VRF) BGP neighbor address-families.

threshold

(Optional) Threshold value (in seconds) used in detecting slow peer. Range is 120 to 3600. Default is 300.

detection-disable

Disables detection-only slow peer configuration on all (default VRF and non-default VRF) BGP neighbor address-families.

Command Default

detection-only

threshold: 300

Command Modes

BGP Router configuration mode

Command History

Release

Modification

Release 7.9.1

This command was introduced.

Usage Guidelines

You can enable slow peer configuration either by using global router configuration mode or by using neighbor address-families.

This section describes global slow peer configuration under router configuration. See below for the effective slow peer configuration when slow peer is configured under both global and neighbor address-families.

-

Global configuration

-

[None]

[Dynamic]

[Detection disable]

Neighbor address-family configuration

[None]

Detection-only

Dynamic

None

[Static]

Static

Static

Static

[Dynamic]

Dynamic

Dynamic

Dynamic

[Dynamic Disable]

Detection-only

None

None

Slow peer global configuration can be configured as dynamic or detection-disable. The default is detection-only.

When global slow peer configuration is not enabled, the default slow peer functionality is detection-only. It means all (default VRF and non-default VRF) BGP neighbor address-families are operating in detection-only mode. In detection-only mode of operation, whenever a neighbor address-family is detected as slow or recovers from being slow, an IOS message is displayed, but there will not be any mitigation to handle slow peers.

When slow peer global configuration is detection-disable, then slow peer processing is disabled on all (default VRF and non-default VRF) BGP neighbor address-families.

When the slow peer global configuration is dynamic, all VRF BGP neighbor address-families, default and non-default, operate as dynamic slow peers. In the dynamic slow peer mode of operation, whenever a neighbor address-family is detected as being slow, in addition to displaying an IOS message, the peer is moved to its own refresh sub-group without impacting other peers in the update-group or a sub-group.

Task ID

Task ID Operation

bgp

read, write

Examples

This example below shows how to enable dynamic slow peer on all (default VRF and non-default VRF) BGP neighbor address-families:
Router#configure
Router(config)#router bgp 100
Router(config-bgp)#slow-peer dynamic
Router(config-bgp)#commit
This example below shows how to disable slow peer on all (default VRF and non-default VRF) BGP neighbor address-families:
Router#configure
Router(config)#router bgp 100
Router(config-bgp)#slow-peer detection-disable
Router(config-bgp)#commit
 
This example below shows how to enable dynamic slow peer with detection threshold of 120 seconds on all (default VRF and non-default VRF) BGP neighbor address-families:
Router#configure
Router(config)#router bgp 100
Router(config-bgp)#slow-peer dynamic threshold 120
Router(config-bgp)#commit 

slow peer (BGP neighbor address-family configuration)

To enable slow peer for a BGP neighbor address-family, use slow-peer command in neighbor address-family configuration mode.

By default, BGP neighbor address-family slow peer configuration is disabled.

Use the no form of the command to return to the default slow peer configuration.

slow-peer {static | dynamic [threshold seconds] | dynamic disable}

Syntax Description

static

Enables static slow peer configuration of a BGP neighbor address-family.

dynamic

Enables dynamic slow peer configuration of a BGP neighbor address-family.

threshold

(Optional) Threshold value (in seconds) used in detecting slow peer. Range is 120 to 3600. Default is 300. The detection logic is evaluated every scanner run whose default interval is 60 seconds.

dynamic disable

Disables dynamic slow peer configuration of a BGP neighbor address-family.

Command Default

threshold: 300

Command Modes

BGP Neighbor address-family configuration

BGP VRF Neighbor address-family configuration

Command History

Release

Modification

Release 7.9.1

This command was introduced.

Usage Guidelines

You can enable slow peer configuration either by using global router configuration mode or by using neighbor address-families.

This section describes slow peer configuration under neighbor address-family configuration. See below for the effective slow peer configuration state when slow peer is configured under both global and neighbor address-family.

-

Global configuration

-

[None]

[Dynamic]

[Detection disable]

Neighbor address-family configuration

[None]

Detection-only

Dynamic

None

[Static]

Static

Static

Static

[Dynamic]

Dynamic

Dynamic

Dynamic

[Dynamic Disable]

Detection-only

None

None

Slow peer configuration under neighbor address-family can be configured as:

  • static

  • dynamic

  • dynamic disable

When slow peer is not configured under neighbor address-family, slow peer handling is disabled for that BGP neighbor address-family.

When slow peer static is configured under a neighbor address-family, then that neighbor address-family is moved into its own unique update-group, thus isolating this neighbor address-family from other neighbors. If the user's intention is to group all the slow-peers into a single update group, it can be accomplished by removing static slow peer configuration and configuring the same neighbor out route-policy for all the neighbors.

When slow peer dynamic is configured under the neighbor address-family, that BGP neighbor address-family is enabled for dynamic slow peer processing. When the neighbor address-family is enabled for dynamic slow peer processing, whenever the neighbor address-family is detected as slow, the neighbor address-family is processed in its own refresh sub-group without affecting other neighbors in the sub-group, in addition to displaying an IOS message indicating the neighbor address-family has become slow.

When slow-peer dynamic disable is configured under the neighbor address-family, it disables the dynamic slow peer processing for that neighbor address-family if dynamic slow peer processing was enabled due to global slow peer dynamic configuration.

Task ID

Task ID Operation

bgp

read, write

Examples

This example below shows how to configure static slow peer for a (default VRF and non-default VRF) BGP neighbor address-family:
Router#configure
Router(config)#router bgp 100
Router(config-bgp)#neighbor 50.0.0.1
Router(config-bgp-nbr)#address-family ipv4 unicast
Router(config-bgp-nbr-af)#slow-peer static
Router(config-bgp-nbr-af)#commit 
This example below shows how to disable slow peer for a (default VRF and non-default VRF) BGP neighbor address-family:
Router#configure
Router(config)#router bgp 100
Router(config-bgp)#neighbor 50.0.0.1
Router(config-bgp-nbr)#address-family ipv4 unicast
Router(config-bgp-nbr-af)#slow-peer dynamic disable
Router(config-bgp-nbr-af)#commit 
This example below shows how to enable dynamic slow peer for a (default VRF and non-default VRF) BGP neighbor address-family:
Router#configure
Router(config)#router bgp 100
Router(config-bgp)#neighbor 50.0.0.1
Router(config-bgp-nbr)#address-family ipv4 unicast
Router(config-bgp-nbr-af)#slow-peer dynamic
Router(config-bgp-nbr-af)#commit 
This example below shows how to enable dynamic slow peer with detection threshold of 120 seconds for a (default VRF and non-default VRF) BGP neighbor address-family:
Router#configure
Router(config)#router bgp 100
Router(config-bgp)#neighbor 50.0.0.1
Router(config-bgp-nbr)#address-family ipv4 unicast
Router(config-bgp-nbr-af)#slow-peer dynamic threshold 120
Router(config-bgp-nbr-af)#commit

socket receive-buffer-size

To set the size of the receive buffers for all Border Gateway Protocol (BGP) neighbors, use the socket receive-buffer-size command in an appropriate configuration mode. To set the size of the receive buffers to the default size, use the no form of this command.

socket receive-buffer-size socket-size [bgp-size]

no socket receive-buffer-size [socket-size] [bgp-size]

Syntax Description

socket-size

Size (in bytes) of the receive-side socket buffers. Range is 512 to 131072.

bgp-size

(Optional) Size (in bytes) of the receive buffers in BGP. Range is 512 to 131072.

Command Default

socket-size : 32,768 bytes

bgp-size : 4,032 bytes

Command Modes

Router configuration

VRF configuration

Command History

Release

Modification

Release 3.7.2

This command was introduced.

Usage Guidelines

To use this command, you must be in a user group associated with a task group that includes appropriate task IDs. If the user group assignment is preventing you from using a command, contact your AAA administrator for assistance.

Use the socket receive-buffer-size command to increase the buffer size when receiving updates from a neighbor. Using larger buffers can improve convergence time because the software can process more packets simultaneously. However, allocating larger buffers uses more memory on your router.


Note


Increasing the socket buffer size uses more memory only when more messages are waiting to be processed by the software. In contrast, increasing the BGP buffer size uses extra memory indefinitely.


Use the receive-buffer-size command on individual neighbors to change the values set by the socket receive-buffer-size command.

Task ID

Task ID

Operations

bgp

read, write

Examples

The following example shows how to set the receive buffer sizes for all neighbors to 65,536 bytes for the socket buffer and 8192 bytes for the BGP buffer:


RP/0/RSP0/CPU0:router(config)# router bgp 1
RP/0/RSP0/CPU0:router(config-bgp)# socket receive-buffer-size 65536 8192

socket send-buffer-size

To set the size of the send buffers for all Border Gateway Protocol (BGP) neighbors, use the socket send-buffer-size command in an appropriate configuration mode. To set the size of the send buffers to the default size, use the no form of this command.

socket send-buffer-size socket-size [bgp-size]

no socket send-buffer-size [socket-size] [bgp-size]

Syntax Description

socket-size

Size (in bytes) of the send-side socket buffers. Range is 4096 to 131072.

bgp-size

(Optional) Size (in bytes) of the send buffers in BGP. Range is 4096 to 131072.

Command Default

socket-size : 10240 bytes

bgp-size : 4096 bytes

Command Modes

Router configuration

VRF configuration

Command History

Release

Modification

Release 3.7.2

This command was introduced.

Usage Guidelines

To use this command, you must be in a user group associated with a task group that includes appropriate task IDs. If the user group assignment is preventing you from using a command, contact your AAA administrator for assistance.

Use the socket send-buffer-size command to increase the buffer size when sending updates to neighbors. Using larger buffers can improve convergence time because the software can process more packets simultaneously. However, allocating larger buffers uses more memory on your router.


Note


Increasing the socket buffer size uses more memory only when more messages are waiting to be sent by the software. In contrast, increasing the BGP buffer size uses extra memory indefinitely.


Use the send-buffer-size command on individual neighbors to change the values set by the socket send-buffer-size command.

Task ID

Task ID

Operations

bgp

read, write

Examples

The following example shows how to set the send buffer sizes for all neighbors to 8192 bytes for the socket buffer and the BGP buffer:


RP/0/RSP0/CPU0:router(config)# router bgp 1
RP/0/RSP0/CPU0:router(config-bgp)# socket send-buffer-size 8192 8192

soft-reconfiguration inbound

To configure the software to store updates received from a neighbor, use the soft-reconfiguration inbound command in an appropriate configuration mode. To disable storing received updates, use the no form of this command.

soft-reconfiguration inbound [ always | inheritance-disable | RPKI-dropped-only | RPKI-tested-only ]

no soft-reconfiguration inbound [ always | inheritance-disable | RPKI-dropped-only | RPKI-tested-only ]

Syntax Description

always

(Optional) Saves a copy of all received routes, even if the neighbor supports the route refresh capability.

inheritance-disable

(Optional) Overrides configuration for this command that may be inherited from a neighbor group or address family group.

RPKI-dropped-only

(Optional) Saves a copy of only the routes dropped by an RPKI validation-state test in neighbor-in route-policy.

RPKI-tested-only

(Optional) Saves a copy of only the routes tested in an RPKI validation-state test in neighbor-in route-policy.

Command Default

Soft reconfiguration is not enabled.

Command Modes

IPv4 address family group configuration

IPv6 address family group configuration

IPv4 neighbor address family configuration

IPv4 neighbor group address family configuration

IPv6 neighbor group address family configuration

VPNv4 neighbor address family configuration

VPNv4 address family group configuration

VRF IPv4 neighbor address family configuration

VPNv4 neighbor group address family configuration

VRF IPv6 neighbor address family configuration

Command History

Release

Modification

Release 3.7.2

This command was introduced.

Release 3.9.0

The disable keyword was replaced with the inheritance-disable keyword.

Usage Guidelines

To use this command, you must be in a user group associated with a task group that includes appropriate task IDs. If the user group assignment is preventing you from using a command, contact your AAA administrator for assistance.

To filter or modify some of the updates received from a neighbor, you configure an inbound policy using the route-policy (BGP) command. Configuring soft reconfiguration inbound causes the software to store the original unmodified route beside a route that is modified or filtered. This allows a “soft clear” to be performed after the inbound policy is changed. To perform a soft clear, use the clear bgp soft command with the in keyword specified. The unmodified routes are then passed through the new policy and installed in the BGP table.


Note


If an address family group, neighbor group, or session group is configured, the configuration inside these configuration groups will not be effective unless it is applied directly or indirectly to one or more neighbors.



Note


The bgp auto-policy-soft-reset is enabled by default. A soft clear is done automatically when the inbound policy configured with the route-policy (BGP) command is changed. This behavior can be changed by disabling the auto-policy-soft-reset using the bgp auto-policy-soft-reset disable command.


If the neighbor supports the route refresh capability, then the original routes are not stored because they can be retrieved from the neighbor through a route refresh request. However, if the always keyword is specified, the original routes are stored even when the neighbor supports the route refresh capability.

If the soft-reconfiguration inbound command is not configured and the neighbor does not support the route refresh capability, then an inbound soft clear is not possible. In that case, the only way to rerun the inbound policy is to use the clear bgp ip-address command to reset the neighbor BGP session.


Note


If there is an existing BGP session with a neighbor that does not support the route refresh capability, the session is terminated and a new one is initiated.



Note


The extra routes stored as a result of configuring this command use more memory on the router.


If you configure this command for a neighbor group or neighbor address family group, all neighbors using the group inherit the configuration. Values of commands configured specifically for a neighbor override inherited values.

Task ID

Task ID

Operations

bgp

read, write

Examples

The following example shows inbound soft reconfiguration enabled for IP Version 4 (IPv4) unicast routes received from neighbor 10.108.1.1. The software stores all routes received in their unmodified form so that when an inbound soft clear is performed later, the stored information can then be used to generate a new set of modified routes.


Router(config)# router bgp 100
Router(config-bgp)# neighbor 10.108.1.1
Router(config-bgp-nbr)# remote-as 100
Router(config-bgp-nbr)# address-family ipv4 unicast
Router(config-bgp-nbr-af)# soft-reconfiguration inbound
Router(config-bgp-nbr-af)# exit

speaker-id

To allocate a speaker process to a neighbor, use the speaker-id command in the appropriate configuration mode. To remove the speaker process from a neighbor, use the no form of this command.

speaker-id id

no speaker-id [id]

Syntax Description

id

ID of the speaker process. Range is 1 to 15.

Command Default

Default is 0.

Command Modes

Neighbor configuration

Session group configuration

Command History

Release

Modification

Release 3.7.2

This command was introduced.

Usage Guidelines

To use this command, you must be in a user group associated with a task group that includes appropriate task IDs. If the user group assignment is preventing you from using a command, contact your AAA administrator for assistance.

Task ID

Task ID

Operations

bgp

read, write

Examples

The following example shows how to allocate speaker process 3 to neighbor 192.168.40.24:


RP/0/RSP0/CPU0:router(config)# router bgp 109
RP/0/RSP0/CPU0:router(config-bgp)# neighbor 192.168.40.24
RP/0/RSP0/CPU0:router(config-bgp-nbr)# speaker-id 3

svd platform enable

To enable selective VRF download (SVD) for Cisco ASR 9000 Series Aggregation Services Router, use the svd platform enable command in administrative configuration mode. To disable selective VRF download for Cisco ASR 9000 Series Aggregation Services Router, use the no form of this command.

svd platform enable

no svd platform enable

Syntax Description

This command has no keywords or arguments.

Usage Guidelines

To use this command, you must be in a user group associated with a task group that includes appropriate task IDs. If the user group assignment is preventing you from using a command, contact your AAA administrator for assistance.

After upgrading to Cisco IOS XR Release 4.3.1 or later, the default setting for SVD is disabled and routes or labels are not selectively downloaded to the line cards. All routes are be available on all line cards. This may lead to out of resource conditions, if the line card hardware scale limits are exceeded because routes were conserved by downloading selectively earlier. To resolve the out of resource condition, turn on SVD using the svd platform enable command.

Ensure that the total number of routes (sum of all unique routes on core facing line cards and edge facing line cards ) fits on the line card hardware, before upgrading to Cisco IOS XR Release 4.3.1 or later.

To enable selective VRF download on Cisco ASR 9000 Series Aggregation Services Router, reload the chassis using the reload location all command after configuring the svd platform enable command.

Note


After enabling SVD using svd platform enable, do not use the selective-vrf-download disable command to turn off SVD.


Selective VRF download is disabled by default. Once SVD is enabled, use the no svd platform enable command followed by a reload of the router using the reload location all command to disable SVD.

Command Default

Selective VRF download is disabled.

Command Modes

Administrative configuration

Command History

Release Modification

Release 4.3.1

This command was introduced.

Usage Guidelines

To use this command, you must be in a user group associated with a task group that includes appropriate task IDs. If the user group assignment is preventing you from using a command, contact your AAA administrator for assistance.

Task ID

Task ID Operation

config-services

read, write

Examples

This example shows how to enable selective VRF download for Cisco ASR 9000 Series Aggregation Services Router:


Router#configure
Router(config)#svd platform enable

table-policy

To apply a routing policy to routes being installed into the routing table, use the table-policy command in an appropriate configuration mode. To disable applying a routing policy when installing routes into the routing table, use the no form of this command.

table-policy policy-name

no table-policy [policy-name]

Syntax Description

policy-name

Name of the routing policy to apply.

Command Default

No policy is applied when routes are installed into the routing table.

Command Modes

IPv4 address family configuration

IPv6 address family configuration

VRF IPv4 address family configuration

VRF IPv6 address family configuration

Command History

Release

Modification

Release 3.7.2

This command was introduced.

Usage Guidelines

To use this command, you must be in a user group associated with a task group that includes appropriate task IDs. If the user group assignment is preventing you from using a command, contact your AAA administrator for assistance.


Note


Table policy provides users with the ability to drop routes from the RIB based on match criteria. This feature can be useful in certain applications and should be used with caution as it can easily create a routing ‘black hole’ where BGP advertises routes to neighbors that BGP does not install in its global routing table and forwarding table.


Use the table-policy command to modify route attributes as the routes are installed into the routing table by Border Gateway Protocol (BGP). Commonly, it is used to set the traffic index attribute.

Task ID

Task ID

Operations

bgp

read, write

Examples

The following example shows how to apply the set-traffic-index policy to IPv4 unicast routes being installed into the routing table:


RP/0/RSP0/CPU0:router(config)# router bgp 1
RP/0/RSP0/CPU0:router(config-bgp)# address-family ipv4 unicast
RP/0/RSP0/CPU0:router(config-bgp-af)# table-policy set-traffic-index

tcp mss inheritance-disable

To disable TCP MSS under neighbor or neighbor group, or to prevent TCP MSS from being inherited from the parent, use the tcp mss inheritance-disable command in the appropriate configuration mode.

tcp mss inheritance-disable

Syntax Description

This command has no keywords or arguments.

Command Default

None

Command Modes

Router configuration mode

Neighbor configuration mode

Neighbor group configuration mode

Command History

Release Modification

Release 5.3.2

This command was introduced.

Usage Guidelines

No specific guidelines impact the use of this command.

Task ID

Task ID Operation

bgp

read, write

Examples

The following example shows how to disable TCP MSS under a specific neighbor:


RP/0/RSP0/CPU0:router#configure
RP/0/RSP0/CPU0:router(config)#router bgp 10
RP/0/RSP0/CPU0:router(config-bgp)#address-family ipv4 unicast
RP/0/RSP0/CPU0:router(config-bgp-af)#exit
RP/0/RSP0/CPU0:router(config-bgp)#neighbor-group n1
RP/0/RSP0/CPU0:router(config-bgp-nbrgrp)#tcp mss 500
RP/0/RSP0/CPU0:router(config-bgp-nbrgrp)#address-family ipv4 unicast
RP/0/RSP0/CPU0:router(config-bgp-nbrgrp-af)#exit
RP/0/RSP0/CPU0:router(config-bgp-nbrgrp)#exit
RP/0/RSP0/CPU0:router(config-bgp)#neighbor 10.0.0.2
RP/0/RSP0/CPU0:router(config-bgp-nbr)#remote-as 1
RP/0/RSP0/CPU0:router(config-bgp-nbr)#use neighbor-group n1
RP/0/RSP0/CPU0:router(config-bgp-nbr)#tcp mss inheritance-disable

timers (BGP)

To set the timers for a specific Border Gateway Protocol (BGP) neighbor, use the timers command in an appropriate configuration mode. To set the timers to the default values, use the no form of this command.

timers keepalive hold-time

no timers [keepalive hold-time]

Syntax Description

keepalive

Frequency (in seconds) with which the software sends keepalive messages to a neighbor. Range is 0 to 65535.

hold-time

Interval (in seconds) after not receiving a keepalive message from the neighbor that the software terminates the BGP session for the neighbor. Values are 0 or a number in the range from 3 to 65535.

Command Default

keepalive : 60 seconds

hold-time : 180 seconds

Use the timers bgp command to override the default values.

Command Modes

Neighbor configuration

VRF neighbor configuration

Neighbor group configuration

Session group configuration

Command History

Release

Modification

Release 3.7.2

This command was introduced.

Usage Guidelines

To use this command, you must be in a user group associated with a task group that includes appropriate task IDs. If the user group assignment is preventing you from using a command, contact your AAA administrator for assistance.

The timers actually used in connection with the neighbor may not be the same as those configured with this command. The actual timers are negotiated with the neighbor when establishing the session. The negotiated hold time is the minimum of the configured time and the hold time received from the neighbor. If the negotiated hold time is 0, keepalives are disabled.

The configured value for the keepalive must not exceed one-third of the negotiated hold time. If it does, a value of one-third of the negotiated hold time is used.

If this command is configured for a neighbor group or neighbor address family group, all neighbors using the group inherit the configuration. Values of commands configured specifically for a neighbor override inherited values.

In cases where mechanisms such as Bi-directional Forwarding Detection (BFD), BGP fast-external-failover or Next-hop Tracking cannot be employed to detect and react to changes in the network in a faster manner, BGP Keepalive and Hold-timer values can be configured to use smaller values than the default (60 and 180 seconds respectively). When using aggressive values, consider the router's profile and scale, particularly in respect to the number of BGP neighbours that will be using sessions with the non-default timers.

Sessions using very aggressive values will be more susceptible to flap during events that cause the Route-Processor's CPU utilization levels to increase. Such events include component OIR, Route-Processor Failover, network instability, excessive churn in routing protocols etc. It is therefore recommended that the desired scale and profile of the router be tested with the non-default timer values, subjecting the router to CPU-intensive events in order to determine the timer threshold values that are appropriate for the router before configuring the values in an operational network.

The BGP Non-Stop Routing (NSR) is able to sustain sessions with more aggressive timer values than BGP Graceful Restart (GR) since in the event of a Route-Processor Failover, Graceful Restart (GR) requires the re-establishment of the TCP session over which the BGP session takes place. When using Non-Stop Routing (NSR), both the underlying TCP session and BGP session are maintained during Route-Processor failover.

Task ID

Task ID

Operations

bgp

read, write

Examples

The following example shows how to change the keepalive timer to 70 seconds and the hold-time timer to 210 seconds for the BGP peer 192.168.40.24:


RP/0/RSP0/CPU0:router(config)# router bgp 109
RP/0/RSP0/CPU0:router(config-bgp)# neighbor 192.168.40.24 
RP/0/RSP0/CPU0:router(config-bgp-nbr)# remote-as 1
RP/0/RSP0/CPU0:router(config-bgp-nbr)# timers 70 210

timers bgp

To change the default timer values for Border Gateway Protocol (BGP) neighbors, use the timers bgp command in an appropriate configuration mode. To set the default timers to the default values, use the no form of this command.

timers bgp keepalive hold-time

no timers bgp [keepalive hold-time]

Syntax Description

keepalive

Frequency (in seconds) with which the software sends keepalive messages to a neighbor. Range is 0 to 65535.

hold-time

Interval (in seconds) after not receiving a keepalive message from the neighbor that the software terminates the BGP session for the neighbor. Values are 0 or a number in the range from 3 to 65535.

Command Default

keepalive : 60 seconds

hold-time : 180 seconds

Command Modes

Router configuration

VRF configuration

Command History

Release

Modification

Release 3.7.2

This command was introduced.

Usage Guidelines

To use this command, you must be in a user group associated with a task group that includes appropriate task IDs. If the user group assignment is preventing you from using a command, contact your AAA administrator for assistance.

Use the timers bgp command to adjust the default timer times used by all BGP neighbors. The values can be overridden on particular neighbors using the timers command in the neighbor configuration mode.

The timers actually used in connection with the neighbor may not be the same as those configured with this command. The actual timers are negotiated with the neighbor when establishing the session. The negotiated hold time is the minimum of the configured time and the hold time received from the neighbor. If the negotiated hold time is 0, keepalives are disabled.

The configured value for the keepalive must not exceed one-third of the negotiated hold time. If it does, a value of one-third of the negotiated hold time is used.

In cases where mechanisms such as Bi-directional Forwarding Detection (BFD), BGP fast-external-failover or Next-hop Tracking cannot be employed to detect and react to changes in the network in a faster manner, BGP Keepalive and Hold-timer values can be configured to use smaller values than the default (60 and 180 seconds respectively). When using aggressive values, consider the router's profile and scale, particularly in respect to the number of BGP neighbors that will be using sessions with the non-default timers.

Sessions using very aggressive values will be more susceptible to flap during events that cause the Route-Processor's CPU utilization levels to increase. Such events include component OIR, Route-Processor Failover, network instability, excessive churn in routing protocols etc. It is therefore recommended that the desired scale and profile of the router be tested with the non-default timer values, subjecting the router to CPU-intensive events in order to determine the timer threshold values that are appropriate for the router before configuring the values in an operational network.

The BGP Non-Stop Routing (NSR) is able to sustain sessions with more aggressive timer values than BGP Graceful Restart (GR) since in the event of a Route-Processor Failover, Graceful Restart (GR) requires the re-establishment of the TCP session over which the BGP session takes place. When using Non-Stop Routing (NSR), both the underlying TCP session and BGP session are maintained during Route-Processor failover.

Task ID

Task ID

Operations

bgp

read, write

Examples

The following example shows how to configure a default keepalive time of 30 seconds and a default hold time of 90 seconds:


RP/0/RSP0/CPU0:router(config)# router bgp 1
RP/0/RSP0/CPU0:router(config-bgp)# timers bgp 30 90

transport (rpki-server)

To choose a transport mechanism for the RPKI cache-server configuration, establish and manage transport connections, and send or receive byte streams from the network, use the transport command in rpki-server configuration mode. To disable the transport connection, use the no form of this command.

transport { ssh | tcp } port port-number

no transport { ssh | tcp } port port-number

Syntax Description

port

Specifies to choose a port number for the RPKI cache transport.

port-number

Specifies the port number for the RPKI cache transport over TCP and SSH protocols. The port number ranges from 1 to 65535.

Note

 

SSH supports custom ports in addition to the default port number 22.

Command Default

Transport mechanism is disabled.

Command Modes

RPKI server configuration

Command History

Release Modification

Release 4.2.1

This command was introduced.

Usage Guidelines

To use this command, you must be in a user group associated with a task group that includes appropriate task IDs. If the user group assignment is preventing you from using a command, contact your AAA administrator for assistance.

The transport can be set to either TCP or SSH. An SSH transport session with port number 22 is the recommended transport between router and RPKI cache for security reasons.

The transport method (TCP or SSH) can be configured on a per-RPKI-server basis: once server can be TCP port 980, another can be SSH port 22, for example. This can be changed by configuration. Changing the transport method will cause the cache session to flap (cleanup its existing transport related data and initialize the new transport related data).

Task ID

Task ID Operation

bgp

read, write

Examples

This example shows how to configure SSH as the transport mechanism and to use port 22 for SSH communication:


Router#configure
Router(config)#router bgp 100
Router(config-bgp)#rpki server 172.168.35.40 
Router(config-bgp-rpki-server)# transport ssh port 22

ttl-security

To configure a router to check the time-to-live (TTL) field in incoming IP packets for the specified external Border Gateway Protocol (eBGP) peer, use the ttl-security command in an appropriate configuration mode. To disable TTL verification, use the no form of this command.

ttl-security [inheritance-disable]

no ttl-security [inheritance-disable]

Syntax Description

inheritance-disable

(Optional) Prevents the ttl-security command from being inherited from a session group or neighbor group.

Command Default

TTL verification is not enabled for eBGP peers.

Command Modes

Neighbor configuration

VRF neighbor configuration

Neighbor group configuration

Session group configuration

Command History

Release

Modification

Release 3.7.2

This command was introduced.

Release 3.9.0

The disable keyword was replaced with the inheritance-disable keyword.

Usage Guidelines

To use this command, you must be in a user group associated with a task group that includes appropriate task IDs. If the user group assignment is preventing you from using a command, contact your AAA administrator for assistance.

Use the ttl-security command to enable a lightweight security mechanism to protect eBGP peering sessions from CPU utilization-based and other resource exhaustion-based attacks. These types of attacks are typically brute-force Denial of Service (DoS) attacks that attempt to disable the network by flooding devices in the network with IP packets that contain forged source and destination IP addresses in the packet headers.

This command leverages existing behavior in IP packets. For a given IP packet, the TTL count of the packet always is equal to or less than the TTL count when the packet originated, a behavior that is considered impossible to circumvent. Therefore, a packet received with a TTL count equal to the maximum TTL value of 255 can be sent only by a directly adjacent peer. When the ttl-security command is configured for an eBGP neighbor that is directly adjacent, the router accepts only IP packets with a TTL count that is equal to the maximum TTL value.

The ttl-security command secures the eBGP session in the incoming direction only. In the outbound direction, it causes packets to be sent only with the maximum TTL value so that the BGP neighbor can also verify the TTL value of incoming packets. When this command is enabled, BGP establishes or maintains a session only if the TTL value in the IP packet header is equal to the maximum TTL value. If the value is less than the maximum TTL value, the packet is discarded and an Internet Control Message Protocol (ICMP) message is not generated. This behavior is designed because a response to a forged packet is not necessary.


Note


The ttl-security command must be configured on each participating router. Failure to configure this command on both ends of the BGP session results in the session progressing as far as the OpenSent or OpenConfirm state, remaining there until the hold time expires.


The following restrictions apply to the configuration of this command:

  • The ttl-security command should not be configured for a peer that is already configured with the neighbor ebgp-multihop command. The simultaneous configuration of these commands is permitted; however, the ttl-security command overrides the ebgp-multihop command.
  • This command is not supported for internal BGP (iBGP) peers.
  • This command is not effective against attacks from a directly adjacent peer that has been compromised.

If you configure this command for a neighbor group or session group, all neighbors using the group inherit the configuration. Values of commands configured specifically for a neighbor override inherited values.


Note


If the ttl-security command is configured on a neighbor to which the router has an established connection or the router is in the process of establishing a connection, the session must be cleared using the clear bgp command.


Task ID

Task ID

Operations

bgp

read, write

Examples

The following example shows how to enable TTL security for eBGP neighbor 192.168.223.7:


RP/0/RSP0/CPU0:router(config)# router bgp 65534
RP/0/RSP0/CPU0:router(config-bgp)# neighbor 192.168.223.7
RP/0/RSP0/CPU0:router(config-bgp-nbr)# remote-as 65507
RP/0/RSP0/CPU0:router(config-bgp-nbr)# ttl-security

The following example shows how to enable TTL security for multiple eBGP neighbors using a session group:


RP/0/RSP0/CPU0:router(config)# router bgp 65534
RP/0/RSP0/CPU0:router(config-bgp)# session-group ebgp-nbrs
RP/0/RSP0/CPU0:router(config-bgp-sngrp)# ttl-security
RP/0/RSP0/CPU0:router(config-bgp-sngrp)# exit
RP/0/RSP0/CPU0:router(config-bgp)# neighbor 192.168.223.1
RP/0/RSP0/CPU0:router(config-bgp-nbr)# remote-as 65501
RP/0/RSP0/CPU0:router(config-bgp-nbr)# use session-group ebgp-nbrs
RP/0/RSP0/CPU0:router(config-bgp-nbr)# exit
RP/0/RSP0/CPU0:router(config-bgp)# neighbor 192.168.223.2
RP/0/RSP0/CPU0:router(config-bgp-nbr)# remote-as 65502
RP/0/RSP0/CPU0:router(config-bgp-nbr)# use session-group ebgp-nbrs
RP/0/RSP0/CPU0:router(config-bgp-nbr)# exit
RP/0/RSP0/CPU0:router(config-bgp)# neighbor 192.168.223.3
RP/0/RSP0/CPU0:router(config-bgp-nbr)# remote-as 65503
RP/0/RSP0/CPU0:router(config-bgp-nbr)# use session-group ebgp-nbrs
RP/0/RSP0/CPU0:router(config-bgp-nbr)# exit

update limit

To set upper bound on transient memory usage for update generation, use the update limit command in router configuration mode. To return the bounds to the default value, use the no form of this command.

update limit update-limit-MB

no update limit

Syntax Description

update-limit-MB

Sets the update limit in megabytes (MB). Range is 16 to 2048 MB.

Command Default

Default update limit is 512 MB.

Command Modes

Router configuration

Command History

Release Modification

Release 4.2.0

This command was introduced and replaced the bgp write-limit command.

Usage Guidelines

To use this command, you must be in a user group associated with a task group that includes appropriate task IDs. If the user group assignment is preventing you from using a command, contact your AAA administrator for assistance.

Use the update limit command to configure a global limit on the size of messages the software queues when updating peers. Increasing the limit can result in faster Border Gateway Protocol (BGP) convergence, but also may result in higher memory usage during convergence.

Task ID

Task ID Operation

bgp

read, write

Examples

This example shows how to set the update limit as 1024 MB:


Router(config)# router bgp 65000
Router(config-bgp)#update limit 1024

update limit address-family

To set upper bound on transient memory usage for update generation for an address family, use the update limit address-family command in an appropriate address-family configuration mode. To return the bounds to the default value, use the no form of this command.

update limit address-family update-limit-MB

no update limit address-family

Syntax Description

update-limit-MB

Sets the update limit in megabytes (MB). Range is 4 MB to 2048 MB.

Command Default

Default update limit is 256 MB.

Command Modes

IPv4 address family configuration

IPv6 address family configuration

L2VPN address family configuration

VPNv4 address family configuration

VPNv6 address family configuration

Command History

Release Modification

Release 4.2.0

This command was introduced and replaced the bgp write-limit command.

Usage Guidelines

To use this command, you must be in a user group associated with a task group that includes appropriate task IDs. If the user group assignment is preventing you from using a command, contact your AAA administrator for assistance.

Use the update limit address-family command to configure a global limit on the size of messages the software queues when updating peers. Increasing the limit can result in faster Border Gateway Protocol (BGP) convergence, but also may result in higher memory usage during convergence.

Task ID

Task ID Operation

bgp

read, write

Examples

This example shows how to set the update limit as 512 MB for address family IPv4 unicast:


Router#configure
Router(config)#router bgp 100
Router(config-bgp)#address-family ipv4 unicast 
Router(config-bgp-af)#update limit address-family 512

update limit sub-group

To set upper bound on transient memory usage for update generation for eBGP or iBGP sub-groups, use the update limit sub-group command in an appropriate address-family configuration mode. To return the bounds to the default value, use the no form of this command.

update limit sub-group { ebgp | ibgp } update-limit-MB

no update limit sub-group { ebgp | ibgp }

Syntax Description

ebgp

Specifies the update limit for eBGP sub-groups.

ibgp

Specifies the update limit for iBGP sub-groups.

update-limit-MB

Sets the update limit in megabytes (MB). Range is 1 MB to 512 MB.

Command Default

Default update limit is 32 MB.

Command Modes

IPv4 address family configuration

IPv6 address family configuration

L2VPN address family configuration

VPNv4 address family configuration

VPNv6 address family configuration

Command History

Release Modification

Release 4.2.0

This command was introduced and replaced the bgp write-limit command.

Usage Guidelines

To use this command, you must be in a user group associated with a task group that includes appropriate task IDs. If the user group assignment is preventing you from using a command, contact your AAA administrator for assistance.

Use the update limit sub-group command to configure a global limit on the size of messages the software queues when updating peers. Increasing the limit can result in faster Border Gateway Protocol (BGP) convergence, but also may result in higher memory usage during convergence.

Task ID

Task ID Operation

bgp

read, write

Examples

This example shows how to set the update limit as 256 MB for eBGP sub-group under address family IPV4 unicast:

Router#configure
Router(config)#router bgp 100
Router(config-bgp)#address-family ipv4 unicast 
Router(config-bgp-af)#update limit sub-group ebgp 256

update in error-handling basic disable

To disable inbound update message basic error handling for eBGP or iBGP neighbors, use the update in error-handling basis disable command in router configuration mode. To enable inbound update message basic error handling, use the no form of this command.

update in error-handling basic {ebgp | ibgp} disable

no update in error-handling basic {ebgp | ibgp} disable

Syntax Description

ebgp

Specifies inbound update message basic error handling for eBGP neighbors.

ibgp

Specifies inbound update message basic error handling for iBGP neighbors.

Command Default

Inbound update message basic error handling is enabled.

Command Modes

Router configuration

Command History

Release Modification

Release 4.2.0

This command was introduced.

Usage Guidelines

To use this command, you must be in a user group associated with a task group that includes appropriate task IDs. If the user group assignment is preventing you from using a command, contact your AAA administrator for assistance.

Task ID

Task ID Operation

bgp

read, write

Examples

This example shows how to disable inbound update message basic error handling for eBGP neighbors:

RP/0/RSP0/CPU0:router#configure
RP/0/RSP0/CPU0:router(config)#router bgp 100
RP/0/RSP0/CPU0:router(config-bgp)#update in error-handling basic ebgp disable

This example shows how to disable inbound update message basic error handling for iBGP neighbors:

RP/0/RSP0/CPU0:router#configure 
RP/0/RSP0/CPU0:router(config)#router bgp 100
RP/0/RSP0/CPU0:router(config-bgp)#update in error-handling basic ibgp disable

update in error-handling extended

To enable inbound update message extended error handling for eBGP or iBGP neighbors, use the update in error-handling extended command in router configuration mode. To disable inbound update message error handling, use the no form of this command.

update in error-handling extended {ebgp | ibgp}

no update in error-handling extended {ebgp | ibgp}

Syntax Description

ebgp

Specifies to enable inbound update message extended error handling for eBGP neighbors.

ibgp

specifies to enable inbound update message extended error handling for iBGP neighbors.

Command Default

Inbound update message extended error handling is disabled.

Command Modes

Router configuration

Command History

Release Modification

Release 4.2.0

This command was introduced.

Usage Guidelines

To use this command, you must be in a user group associated with a task group that includes appropriate task IDs. If the user group assignment is preventing you from using a command, contact your AAA administrator for assistance.

Task ID

Task ID Operation

bgp

read, write

Examples

This example shows how to enable inbound update message extended error handling for eBGP neighbors:

RP/0/RSP0/CPU0:router#configure 
RP/0/RSP0/CPU0:router(config)#router bgp 100
RP/0/RSP0/CPU0:router(config-bgp)#update in error-handling extended ebgp

This example shows how to enable inbound update message extended error handling for iBGP neighbors:

RP/0/RSP0/CPU0:router#configure 
RP/0/RSP0/CPU0:router(config)#router bgp 100
RP/0/RSP0/CPU0:router(config-bgp)#update in error-handling extended ibgp

update in error-handling treat-as-withdraw (BGP)

To discard the incoming update message as a withdraw message and to avoid BGP session reset, use the update in error-handling treat-as-withdraw command in Global Configuration mode. To disable withdraw during inbound update message error handling, use the no form of this command.

update in error-handling treat-as-withdraw

Syntax Description

This command has no keywords or arguments.

Command Modes

Router configuration

Command History

Release

Modification

Release 7.10.1

This command was introduced.

Usage Guidelines

No specific guidelines impact the use of this command.

Task ID

Task ID Operation

bgp

read, write

Examples

This example shows how to discard the incoming update message as a withdraw message and to avoid BGP session reset:
Router(config)#router bgp 109 
Router(config-bgp)#neighbor 192.168.40.24 
Router(config-bgp-nbr)#update in error-handling treat-as-withdraw
Router(config-bgp-nbr)#commit

update in labeled-unicast equivalent

To treat SAFIs unicast and labelled unicast for a given neighbor as same, use the update in labeled-unicast equivalent command in neighbor configuration mode. This command was introduced for interoperability between Cisco IOS XR and IOS XE systems. For example, If the XE router advertises ipv4 unicast and then withdraw with the ipv4 label-unicast of the same route, with knob enabled, then the BGP in the receiving XR router treats both of them same. Hence BGP removes the prefix information from its database.

update in labeled-unicast equivalent [inheritance-disable]

Syntax Description

equivalent

IPv4/IPv6 labeled-unicast inbound updates (paths) treated equivalent to unicast updates (paths)

inheritance-disable

Does not inherit this unicast configuration from parent group

Command Default

None

Command Modes

Neighbor configuration mode

Neighbor group configuration mode

Command History

Release Modification

Release 7.0.1

This command was introduced.

Usage Guidelines

To use this command, you must be in a user group associated with a task group that includes appropriate task IDs. If the user group assignment is preventing you from using a command, contact your AAA administrator for assistance.

Task ID

Task ID Operation

bgp

read, write

Examples

The following example shows how to treat IPv4 labeled-unicast inbound updates equivalent to unicast updates (paths):


RP/0/RSP0/CPU0:router#configure
RP/0/RSP0/CPU0:router(config)#router bgp 100
RP/0/RSP0/CPU0:router(config-bgp)#neighbor 10.2.3.4
RP/0/RSP0/CPU0:router(config-bgp-nbr)#update in labeled-unicast equivalentexit

update out logging

To enable logging of update generation events, use the update out logging command in router configuration mode. To disable the logging of update generation events, use the no form of this command.

update out logging

no update out logging

Syntax Description

This command has no keywords or arguments.

Command Default

Update generation event logging is disabled.

Command Modes

Router configuration

Command History

Release Modification

Release 4.2.0

This command was introduced.

Usage Guidelines

To use this command, you must be in a user group associated with a task group that includes appropriate task IDs. If the user group assignment is preventing you from using a command, contact your AAA administrator for assistance.

Task ID

Task ID Operation

bgp

read, write

Examples

This example shows how to enable logging of update generation events:

RP/0/RSP0/CPU0:router#configure 
RP/0/RSP0/CPU0:router(config)#router bgp 100
RP/0/RSP0/CPU0:router(config-bgp)#update out logging

update out originator-loopcheck disable

To bypass originator loop check for outgoing update, use the update out originator-loopcheck disable command in neighbor adress-family configuration mode. This configuration is effective only for the neighbor in unique update group for which the update is being processed.

update out originator-loopcheck [disable]

To restore the system to its default condition, use the no form of this command or use the update out originator-loopcheck command.

no update out originator-loopcheck [disable]

update out originator-loopcheck

Syntax Description

disable (optional)

Prevents the update out originator-loopcheck disable command from being inherited from a parent group.

Command Default

Originator loop check for outbound updates is enabled if there is only one neighbor, and disabled if there are multiple neighbors in the update group.

Command Modes

Non-VRF neighbor address family configuration

Supported Address-families

  • address-family vpnv4 unicast

  • address-family vpnv6 unicast

  • address-family l2vpn evpn

Command History

Release Modification
Release 6.6.2 This command was introduced.

Usage Guidelines

Use the update out originator-loopcheck disable command to bypass originator loop check for outgoing update. Use this command either with as-override command or with as-path-loopcheck out disable command for eBGP peering. For iBGP peering, separate configuration is not required.

The word “originator” represents the either of the following:

  • The neighbor whose IP address matches with the current network path's neighbor IP address and the update is sent to that same neighbor.

  • The neighbor for the route reflector or iBGP, when the neighbor's router-id matches with the current network path's router-id, and the update is sent to that same neighbor.

Although, the disable keyword is optional, you must configure it to disable the originator loop check. For example, under neighbor address family, if the update out originator-loopcheck disable command is configured, only then it disables the originator loop check for that neighbor. There can be scenarios where configuring the command without the optional disable keyword is useful.

Task ID

Task ID Operation

BGP

read, write

Examples

Example 1:

RP/0/RSP0/CPU0:router#configure
RP/0/RSP0/CPU0:router(config)#router bgp 100
RP/0/RSP0/CPU0:router(config-bgp)# neighbor 10.40.0.1
RP/0/RSP0/CPU0:router(config-bgp)# remote-as 200
RP/0/RSP0/CPU0:router (config-bgp-nbr)#address-family l2vpn evpn 
RP/0/RSP0/CPU0:router (config-bgp-nbr-af)#update out originator-loopcheck disable
RP/0/RSP0/CPU0:router(config-bgp-nbr-af)#as-override 

Examples

Example 2:

RP/0/RSP0/CPU0:router#configure
RP/0/RSP0/CPU0:router(config)#router bgp 100
RP/0/RSP0/CPU0:router(config-bgp)#address-family l2vpn evpn
RP/0/RSP0/CPU0:router(config-bgp-af)#as-path-loopcheck out disable
RP/0/RSP0/CPU0:router(config-bgp)#neighbor 10.40.0.1
RP/0/RSP0/CPU0:router(config-bgp-nbr)#remote-as 200
RP/0/RSP0/CPU0:router (config-bgp-nbr)#address-family l2vpn evpn
RP/0/RSP0/CPU0:router (config-bgp-nbr-af)#update out originator-loopcheck disable

Examples

Example 3:

RP/0/RSP0/CPU0:router#configure
RP/0/RSP0/CPU0:router(config)#router bgp 100
RP/0/RSP0/CPU0:router(config-bgp)#neighbor 10.60.0.1
RP/0/RSP0/CPU0:router(config-bgp-nbr)#remote-as 100
RP/0/RSP0/CPU0:router (config-bgp-nbr)#address-family vpnv4 unicast
RP/0/RSP0/CPU0:router (config-bgp-nbr-af)#update out originator-loopcheck disable

update-source

To allow internal Border Gateway Protocol (iBGP) sessions to use the primary IP address from a particular interface as the local address when forming an iBGP session with a neighbor, use the update-source command in an appropriate configuration mode. To set the chosen local IP address to the nearest interface to the neighbor, use the no form of this command.

update-source type interface-path-id

no update-source [type interface-path-id]

Syntax Description

type

Interface type. For more information, use the question mark ( ? ) online help function.

interface-path-id

Physical interface or virtual interface.

Note

 

Use the show interfaces command to see a list of all interfaces currently configured on the router.

For more information about the syntax for the router, use the question mark ( ? ) online help function.

Command Default

Best local address

Command Modes

Neighbor configuration

VRF neighbor configuration

Neighbor group configuration

Session group configuration

Command History

Release

Modification

Release 3.7.2

This command was introduced.

Usage Guidelines

To use this command, you must be in a user group associated with a task group that includes appropriate task IDs. If the user group assignment is preventing you from using a command, contact your AAA administrator for assistance.

The update-source command is commonly used with the loopback interface feature for iBGP sessions. The loopback interface is defined, and the interface address is used as the endpoint for a BGP session through the update-source command. This mechanism allows a BGP session to remain up even if the outbound interface goes down, provided there is another route to the neighbor.

If this command is configured for a neighbor group or session group, all neighbors using the group inherit the configuration. Values of commands configured specifically for a neighbor override inherited values.

Task ID

Task ID

Operations

bgp

read, write

Examples

The following example shows how to configure this router to use the IP address from the Loopback0 interface when trying to open a session with neighbor 172.20.16.6:


RP/0/RSP0/CPU0:router(config)# router bgp 110
RP/0/RSP0/CPU0:router(config-bgp)# neighbor 172.20.16.6
RP/0/RSP0/CPU0:router(config-bgp-nbr)# remote-as 110
RP/0/RSP0/CPU0:router(config-bgp-nbr)# update-source Loopback0

update wait-install

To configure BGP to wait for feedback from RIB indicating that the routes that BGP installed in RIB have been installed in FIB, before BGP send out updates to neighbors, use the update wait-install command in an appropriate configuration mode.

update wait-install

no update wait-install

This command has no keywords or arguments.

Command Default

The update wait-install configuration is disabled.

Command Modes

Router IPv4 address family

Router VPNv4 address family

Command History

Release Modification
Release 4.3.0

This command was introduced.

Usage Guidelines

To use this command, you must be in a user group associated with a task group that includes appropriate task IDs. If the user group assignment is preventing you from using a command, contact your AAA administrator for assistance.

Task ID

Task ID Operation

bgp

read, write

Examples

This example shows how to enable the update wait-install configuration under VPNv4 unicast address family:


Router#configure
Router(config)#router bgp 65500
Router(config-bgp)#address-family vpnv4 unicast
Router(config-bgp-af)#update wait-install

update wait-install delay startup

To schedule the delay for BGP update generations to prevent routes from being advertised to peers until RIB is synchronized, run the update wait-install delay startup command in the System Admin Config mode.

Use the no form of this command to disable the feature.

update wait-install delay startup seconds

Syntax Description

seconds

Delay for the BGP start-up phase. The range of the delay is from 1 second to 600 seconds.

Command Default

The command is disabled by default.

Command Modes

System Admin Config mode

Command History

Release

Modification

Release 7.5.3

This command was introduced.

Usage Guidelines

This command is applicable for the following Address Family Indicators (AFIs):

  • IPv4 unicast

  • IPv6 unicast

  • VPNv4 unicast

  • VPNv6 unicast

Task ID

Task ID

Operations

bgp

read, write

Examples

The following example shows how to schedule the delay of the BGP update generation in the start-up phase to prevent routes from being advertised to peers until RIB is synchronized.

Router# configure
Router(config)# router bgp 1
Router(config-bgp)# address-family ipv4 unicast
Router(config-bgp-af)# update wait-install delay startup 10 
Router(config-bgp-af)# commit

use

To inherit configuration from a neighbor group, session group, or address family group, use the use command in an appropriate configuration mode. To discontinue inheritance from a group, use the no form of this command.

use {af-group group-name | neighbor-group group-name | session-group group-nam e}

no use {af-group [group-name] | neighbor-group [group-name] | session-group [group-name]}

Syntax Description

af-group

Specifies an address family group.

group-name

Name of the neighbor group, session group, or address family group from which you want to inherit configuration.

neighbor-group

Specifies a neighbor group.

session-group

Specifies a session group.

Command Default

Inheritance of group characteristics does not occur.

Command Modes

For use af-group version:

Address family group configuration

Neighbor address family configuration

Neighbor group address family configuration

For use neighbor-group version:

Neighbor group configuration

Neighbor configuration

VRF neighbor configuration

For use session-group version:

Neighbor group configuration

Neighbor configuration

VRF neighbor configuration

Session-group configuration

Command History

Release

Modification

Release 3.7.2

This command was introduced.

Usage Guidelines

To use this command, you must be in a user group associated with a task group that includes appropriate task IDs. If the user group assignment is preventing you from using a command, contact your AAA administrator for assistance.

The use command configures inheritance of configuration from an address family group, neighbor group, or session group, which means that any configuration for the group also takes effect for the user of the group.

The configuration inherited depends on the type of group that is specified. The group types are described in the following sections:

Address Family Group

An address family group can specify a configuration for only a single address family. The address family specified when the address family group was defined (through the af-group command) must match the address family from which the group is used.

Neighbor Group

A neighbor group (like a neighbor) can have address family-independent configuration and address family-specific configuration. All of these configurations could be inherited.

Session Group

A session group can have only address family-independent configuration and thus only address family-independent configuration is inherited from it.

The following rules govern inheritance to resolve possible conflicting configuration:

  1. If a command is configured directly on the neighbor that is using group configuration, the command overrides the value that would be normally inherited from the group.
  2. If the neighbor is configured to use a session group (for address family-independent configuration) or an address family group (for address family-specific configuration) and the command is configured for the session group or address family group, that configuration is used.
  3. The neighbor group configuration is used:

    • If the command is not configured directly on the neighbor and the neighbor is not using a session group (for address family-independent configuration) or an af-group (for address family-specific configuration).

    • The neighbor is using a neighbor group and the command is configured on the neighbor group.

Typically, all configuration for a neighbor group is inherited, but some characteristics may be masked by a session group or address family group. For an example of this configuration, see the “Examples” section.

If the neighbor is using both a session group and a neighbor group and a specific command is configured for the neighbor group but not for the session group, then the configuration for the neighbor group does not take effect. The session group “hides” all address family-independent configuration on the neighbor group and prevents it from being inherited. Similarly, the use of an address family group hides any address family-specific configuration that may otherwise be inherited from a neighbor group for that address family.

In addition to neighbors using groups, it is possible to build a hierarchy by having groups use other groups. The following hierarchical groups are permitted:

  • Session groups may use other session groups.
  • Address family groups may use other address family groups.
  • Neighbor groups may use other neighbor groups.
  • Neighbor groups may use session groups and address family groups.

Note


Within the Cisco IOS XR system configuration architecture, do not combine the remote-as command and the no use neighbor-group command in the same commit, or the remote-as command and the no use session-group command in the same commit.


Task ID

Task ID

Operations

bgp

read, write

Examples

The following example shows how to define a session group session1 and configure neighbor 172.168.40.24 to use session1. As a result, the session1 configuration takes effect on the neighbor also.


Router(config)# router bgp 1
Router(config-bgp)# session-group session1
Router(config-bgp-sngrp)# advertisement-interval 40
Router(config-bgp-sngrp)# timers 30 90
Router(config-bgp-sngrp)# exit
Router(config-bgp)# neighbor 172.168.40.24
Router(config-bgp-nbr)# remote-as 2
Router(config-bgp-nbr)# use session-group session1
Router(config-bgp-nbr)# exit

The following example is similar to the previous example, but in this case the timers command on the session group does not take effect on the neighbor because it is overridden by a timers command directly configured for the neighbor.


Router(config)# router bgp 1
Router(config-bgp)# session-group session1
Router(config-bgp-sngrp)# advertisement-interval 40
Router(config-bgp-sngrp)# timers 30 90
Router(config-bgp-sngrp)# exit
Router(config-bgp)# neighbor 172.168.40.24
Router(config-bgp-nbr)# remote-as 2
Router(config-bgp-nbr)# use session-group session1
Router(config-bgp-nbr)# timers 60 180
Router(config-bgp-nbr)# exit

The following example shows an address family group, family1, for IPv4 multicast and a neighbor group, neighbor1, that have IPv4 unicast and IPv4 multicast enabled. In this case, the neighbor inherits IPv4 unicast (and address family-independent) configuration from the neighbor group, but inherits IPv4 multicast configuration from the address family group. In this example, the neighbor group also has a remote autonomous system configured, so there is no need to configure a remote autonomous system for the neighbor because it inherits the remote autonomous system from the neighbor group:

Router(config)# router bgp 1
Router(config-bgp)# af-group family1 address-family ipv4 multicast
Router(config-bgp-afgrp)# route-policy mcast-in in
Router(config-bgp-afgrp)# exit
Router(config-bgp)# neighbor-group neighbor1
Router(config-bgp-nbrgrp)# remote-as 2
Router(config-bgp-nbrgrp)# address-family ipv4 unicast
Router(config-bgp-nbrgrp-af)# route-policy policy1 in
Router(config-bgp-nbrgrp-af)# route-policy policy1 out
Router(config-bgp-nbrgrp-af)# exit
Router(config-bgp-nbrgrp)# address-family ipv4 multicast
Router(config-bgp-nbrgrp-af)# route-policy policy1 in
Router(config-bgp-nbrgrp-af)# route-policy policy1 out
Router(config-bgp-nbrgrp-af)# exit
Router(config-bgp)# neighbor 172.168.40.24
Router(config-bgp-nbr)# use neighbor-group neighbor1
Router(config-bgp-nbr)# address-family ipv4 multicast
Router(config-bgp-nbr-af)# use af-group family1
Router(config-bgp-nbr-af)# exit

In the previous example, the neighbor uses the policy1 route policy for inbound and outbound IPv4 unicast routes, but uses the mcast-in route policy for inbound IPv4 multicast routes and no policy for outbound IPv4 multicast routes.

The following example shows a neighbor inheriting configuration from a session group that likewise inherits configuration from another session group. The configuration from both session groups take effect on the neighbor:


Router(config)# router bgp 1
Router(config-bgp)# session-group session1
Router(config-bgp-sngrp)# advertisement-interval 40
Router(config-bgp-sngrp)# exit
Router(config-bgp)# session-group session2
Router(config-bgp-sngrp)# use session-group session1
Router(config-bgp-sngrp)# update-source Loopback0
Router(config-bgp-sngrp)# exit
Router(config-bgp)# neighbor 172.168.40.24
Router(config-bgp-nbr)# remote-as 1
Router(config-bgp-nbr)# use session-group session2
Router(config-bgp-nbr)# exit

username (rpki-server)

To specify a SSH username for the RPKI cache-server, use the username command in rpki-server configuration mode. To remove the username, use the no form of this command.

username user-name

no username user-name

Syntax Description

user-name

Enters a username to be used for the SSH transport mechanism.

Command Default

Username is not configured.

Command Modes

RPKI server configuration

Command History

Release Modification

Release 4.2.1

This command was introduced.

Usage Guidelines

To use this command, you must be in a user group associated with a task group that includes appropriate task IDs. If the user group assignment is preventing you from using a command, contact your AAA administrator for assistance.

The username configuration applies only if the SSH transport mechanism is active.

Task ID

Task ID Operation

bgp

read, write

Examples

This example shows how to configure a username (rpki-user ) for the RPKI cache-server SSH transport mechanism:


RP/0/RSP0/CPU0:router#configure
RP/0/RSP0/CPU0:router(config)#router bgp 100
RP/0/RSP0/CPU0:router(config-bgp)#rpki server 172.168.35.40 
RP/0/RSP0/CPU0:router(config-bgp-rpki-server)# transport ssh port 22
RP/0/RSP0/CPU0:router(config-bgp-rpki-server)#username rpki-user

vrf (BGP)

To configure a VPN routing and forwarding (VRF) instance and enter VRF configuration mode, use the vrf command in router configuration mode. To remove the VRF instance from the configuration file and restore the system to its default condition, use the no form of this command.

vrf vrf-name

no vrf vrf-name

Syntax Description

vrf-name

Name of the VRF instance. The following names cannot be used: all, default, and global.

Command Default

No default behavior or values

Command Modes

Router configuration

Command History

Release

Modification

Release 3.7.2

This command was introduced.

Usage Guidelines

To use this command, you must be in a user group associated with a task group that includes appropriate task IDs. If the user group assignment is preventing you from using a command, contact your AAA administrator for assistance.

Use the vrf command to configure a VRF instance. A VRF instance is a collection of VPN routing and forwarding tables maintained at the provider edge (PE) router.

Task ID

Task ID

Operations

bgp

read, write

Examples

The following example shows how to configure a VRF instance and enter VRF configuration mode:


RP/0/RSP0/CPU0:router(config)# router bgp 1
RP/0/RSP0/CPU0:router(config-bgp)# vrf vrf-1
RP/0/RSP0/CPU0:router(config-bgp-vrf)# 

weight

To assign a weight to routes received from a neighbor, use the weight command in an appropriate configuration mode. To remove the weight command from the configuration file and restore the system to its default condition in which the software assigns the default weight to routes, use the no form of this command.

weight weight-value

no weight [weight-value]

Syntax Description

weight-value

Weight to assign. Range is 0 to 65535.

Command Default

Routes learned through another Border Gateway Protocol (BGP) peer have a default weight of 0 and routes sourced by the local router have a default weight of 32768.

Command Modes

IPv4 address family group configuration

IPv6 address family group configuration

IPv4 neighbor address family configuration

IPv4 neighbor group address family configuration

IPv6 neighbor group address family configuration

VPNv4 address family group configuration

VPNv4 neighbor address family configuration

VRF IPv4 neighbor address family configuration

VPNv4 neighbor group address family configuration

VRF IPv6 neighbor address family configuration

Command History

Release

Modification

Release 3.7.2

This command was introduced.

Usage Guidelines

To use this command, you must be in a user group associated with a task group that includes appropriate task IDs. If the user group assignment is preventing you from using a command, contact your AAA administrator for assistance.

The weight of a route is a Cisco-specific attribute. It is used in the best-path selection process (as the strongest tie-breaker). See the Implementing BGP on Cisco ASR 9000 Series Router module of the Routing Configuration Guide for Cisco ASR 9000 Series Routers for information on best path. If there are two BGP routes with the same network layer reachability information (NLRI), the route with the higher weight is always chosen no matter what the value of other BGP attributes. Weight only has significance on the local router. Weight is assigned locally to the router, is a value that only makes sense to the specific router, is not propagated or carried through any route updates, and never is sent between BGP peers (even within the same AS).

Note


If an address family group, neighbor group, or session group is configured, the configuration inside these configuration groups will not be effective unless it is applied directly or indirectly to one or more neighbors.


The weight assigned to individual routes can be further manipulated in the inbound route policy of a neighbor using the set weight command. The set weight command sets the weight directly. If you have particular neighbors that you want to prefer for most of your outbound traffic, you can assign a higher weight to all routes learned from that neighbor.

The weight assigned to individual routes may be modified by using an inbound routing policy.


Note


For weight changes to take effect, you may need to use the clear bgp soft command.


If this command configures a neighbor group or neighbor address family group, all neighbors using the group inherit the configuration. Values of commands configured specifically for a neighbor override inherited values.

Task ID

Task ID

Operations

bgp

read, write

Examples

The following example shows how to assign a weight of 50 to all IP Version 4 (IPv4) unicast routes learned through 172.20.16.6:


RP/0/RSP0/CPU0:router(config)# router bgp 1
RP/0/RSP0/CPU0:router(config-bgp)# neighbor 172.20.16.6
RP/0/RSP0/CPU0:router(config-bgp-nbr)# remote-as 1
RP/0/RSP0/CPU0:router(config-bgp-nbr)# address-family ipv4 unicast
RP/0/RSP0/CPU0:router(config-bgp-nbr-af)# weight 50
RP/0/RSP0/CPU0:router(config-bgp-nbr-af)# exit