Catalyst 4500 Series Switch Software Configuration Guide, Release IOS XE 3.3(0)XO

Logical Layer 3 VLAN Interfaces

The logical Layer 3 VLAN interfaces provide logical routing interfaces to VLANs on Layer 2 switches. A traditional network requires a physical interface from a router to a switch to perform inter-VLAN routing. The Catalyst 4500 series switch supports inter-VLAN routing by integrating the routing and bridging functions on a single Catalyst 4500 series switch.

Figure 30-1 shows how the routing and bridging functions in the three physical devices of the traditional network are performed logically on one Catalyst 4500 series switch.

Figure 30-1 Logical Layer 3 VLAN Interfaces for the Catalyst 4500 Series Switch

Physical Layer 3 Interfaces

The physical Layer 3 interfaces support capabilities equivalent to a traditional router. These Layer 3 interfaces provide hosts with physical routing interfaces to a Catalyst 4500 series switch.

Figure 30-2 shows how the Catalyst 4500 series switch functions as a traditional router.

Figure 30-2 Physical Layer 3 Interfaces for the Catalyst 4500 Series Switch

Understanding SVI Autostate Exclude

To be up/up, a router VLAN interface must fulfill the following general conditions:

• The VLAN exists and is active on the VLAN database of the switch.
• The VLAN interface exists on the router and is not administratively down.
• At least one Layer 2 (access port or trunk) port exists, has a link up on this VLAN, and is in spanning-tree forwarding state on the VLAN.

Note The protocol line state for the VLAN interfaces comes up when the first switch port belonging to the corresponding VLAN link comes up and is in spanning-tree forwarding state.

Ordinarily, when a VLAN interface has multiple ports in the VLAN, the SVI goes down when all the ports in the VLAN go down. The SVI Autostate Exclude feature provides a knob to mark a port so that it is not counted in the SVI up and down calculation and applies to all VLANs that are enabled on that port.

A VLAN interface is brought up after the Layer 2 port has had time to converge (that is, transition from listening-learning to forwarding). This prevents routing protocols and other features from using the VLAN interface as if it were fully operational. It also prevents other problems from occurring, such as routing black holes.

Understanding Layer 3 Interface Counters

Note Supervisor Engine 8-E does not support Layer 2 interface counters. However, it does support Layer 3 (SVI) interface counters.

When you run IPv4 and IPv6 on Supervisor Engine 8-E packets are routed in hardware by the forwarding engine. They support the following statistics for counting routed packets with a maximum of 4092 interfaces:

• Input unicast
• Input multicast
• Output unicast
• Output multicast

For each counter type, both the number of packets and the total number of bytes received or transmitted are counted. You can collect these statistics uniquely for IPv4 and IPv6 traffic.

Because the total number of supported Layer 3 interfaces exceeds the number of counters supported by hardware, all Layer 3 interfaces might not have counters. You assign counters to Layer 3 interfaces; the default configuration for a Layer 3 interface has no counters.

You can configure collection statistics at an interface level in one of the four ways (see Table 30-1 ). The maximum number of interfaces applied to the configuration depends on the collection mode.

When mixing these configured modes, the rule is as follows:

(number of v4/v6/v4v6combined interfaces) + 2*(number of v4v6separate interfaces) <= 4092

Note To enable Layer 3 interface counters, you need to enter the counter command in interface mode. For instructions, see the “Configuring Layer 3 Interface Counters” section.

The hardware counters are displayed in the output of the show interface command, as shown in the following example. Counter fields that are updated when the counter configuration is present are highlighted.

The output of the previous configuration depends on the counter configuration ( Table 30-2 ).

Configuring SVI Autostate Exclude

Note The SVI Autostate Exclude feature is enabled by default and is synchronized with the STP state.

The SVI Autostate Exclude feature shuts down (or brings up) the Layer 3 interfaces of a switch when the following port configuration changes occur:

• When the last port on a VLAN goes down, the Layer 3 interface on that VLAN is shut down
  (SVI- autostated).
• When the first port on the VLAN is brought back up, the Layer 3 interface on the VLAN that was previously shut down is brought up.

SVI Autostate Exclude enables you to exclude the access ports and trunks in defining the status of the SVI (up or down) even if it belongs to the same VLAN. If the excluded access port and trunk is in up state and other ports are in down state in the VLAN, the SVI state is changed to down.

To make the SVI state up, at least one port in the VLAN should be up and not excluded. This action helps to exclude the monitoring port status when you are determining the status of the SVI.

To apply SVI Autostate Exclude, perform this task:

This example shows how to apply SVI Autostate Exclude on interface g3/1:

Configuring IP MTU Sizes

You can set the protocol-specific maximum transmission unit (MTU) size of IPv4 or IPv6 packets that are sent on an interface.

For information on MTU limitations, refer to “Maximum Transmission Units” section.

Note To set the nonprotocol-specific MTU value for an interface, use the mtu interface configuration command. Changing the MTU value (with the mtu interface configuration command) can affect the IP MTU value. If the current IP MTU value matches the MTU value, and you change the MTU value, the IP MTU value is modified automatically to match the new MTU. However, the reverse is not true; changing the IP MTU value has no effect on the value for the mtu command.

For information on how to configure MTU size, refer to “Configuring MTU Sizes” section.

To set the protocol-specific maximum transmission unit (MTU) size of IPv4 or IPv6 packets sent on an interface, perform this task:

This example shows how to configure IPv4 MTU on an interface:

The following example shows how to configure IPv6 MTU on an interface:

This example shows how to verify the configuration

Note When IPv6 is enabled on an interface using any CLI command, you may see the following message:

% Hardware MTU table exhausted

In this situation, the IPv6 MTU value programmed in hardware differs from the IPv6 interface MTU value. This situation occurs if no room exists in the hardware MTU table to store additional values. You must free up some space in the table by unconfiguring some unused MTU values and subsequently disable and reenable IPv6 on the interface or reapply the MTU configuration.

Configuring Layer 3 Interface Counters

Note Supervisor Engine 8-E does not support Layer 2 interface counters.

To configure Layer 3 interface counters (assign counters to a Layer 3 interface), perform this task:

This example shows how to enable counters on interface VLAN 1:

Note To remove the counters, use the no counter command.

If you have already assigned the maximum number of counters, the counter command fails and displays an error message:

In this situation, you must release a counter from another interface for use by the new interface.

About EIGRP Stub Routing

The EIGRP stub routing feature, available in all images, reduces resource utilization by moving routed traffic closer to the end user.

The IP base image contains only EIGRP stub routing. The IP services image contains complete EIGRP routing.

In a network using EIGRP stub routing, the only route for IP traffic to follow to the user is through a switch that is configured with EIGRP stub routing. The switch sends the routed traffic to interfaces that are configured as user interfaces or are connected to other devices.

When using EIGRP stub routing, you need to configure the distribution and remote switches to use EIGRP, and to configure only the switch as a stub. Only specified routes are propagated from the switch. The switch responds to all queries for summaries, connected routes, and routing updates.

Any neighbor that receives a packet informing it of the stub status does not query the stub switch for any routes, and a switch that has a stub peer does not query that peer. The stub switch depends on the distribution switch to send the proper updates to all peers.

In Figure 30-3, switch B is configured as an EIGRP stub switch. Switches A and C are connected to the rest of the WAN. Switch B advertises connected, static, redistribution, and summary routes from switch A and C to Hosts A, B, and C. Switch B does not advertise any routes learned from switch A (and the reverse).

Figure 30-3 EIGRP Stub Switch Configuration

For more information about EIGRP stub routing, see the “Configuring EIGRP Stub Routing” part of the Cisco IOS IP Configuration Guide, Volume 2 of 3: Routing Protocols, Release 12.2.

Configuring EIGRP Stub Routing

The EIGRP stub routing feature improves network stability, reduces resource utilization, and simplifies stub switch configuration.

Stub routing is commonly used in a hub-and-spoke network topology. In a hub-and-spoke network, one or more end (stub) networks are connected to a remote switch (the spoke) that is connected to one or more distribution switches (the hub). The remote switch is adjacent only to one or more distribution switches. The only route for IP traffic to follow into the remote switch is through a distribution switch. This type of configuration is commonly used in WAN topologies where the distribution switch is directly connected to a WAN. The distribution switch can be connected to many more remote switches. Often, the distribution switch is connected to 100 or more remote routers. In a hub-and-spoke topology, the remote router must forward all nonlocal traffic to a distribution router, so it becomes unnecessary for the remote router to hold a complete routing table. Generally, the distribution router need not send anything more than a default route to the remote router.

When using the EIGRP stub routing feature, you need to configure the distribution and remote routers to use EIGRP, and to configure only the remote router as a stub. Only specified routes are propagated from the remote (stub) router. The stub router responds to all queries for summaries, connected routes, redistributed static routes, external routes, and internal routes with the message “inaccessible.” A router that is configured as a stub sends a special peer information packet to all neighboring routers to report its status as a stub router.

Any neighbor that receives a packet informing it of the stub status does not query the stub router for any routes, and a router that has a stub peer does not query that peer. The stub router depends on the distribution router to send the proper updates to all peers.

Figure 30-4 shows a simple hub-and-spoke configuration.

Figure 30-4 Simple Hub-and-Spoke Network

The stub routing feature does not prevent routes from being advertised to the remote router. In the example in Figure 30-4, the remote router can access the corporate network and the Internet using a distribution router only. In this example, having a full route table on the remote router serves no purpose because the path to the corporate network and the Internet always uses a distribution router. The larger route table only reduces the amount of memory required by the remote router. Bandwidth and memory can be conserved by summarizing and filtering routes in the distribution router. The remote router need not receive routes that have been learned from other networks because the remote router must send all nonlocal traffic, regardless of destination, to the distribution router. If a true stub network is desired, the distribution router should be configured to send only a default route to the remote router. The EIGRP stub routing feature does not automatically enable summarization on the distribution router. In most cases, the network administrator needs to configure summarization on the distribution routers.

Note When configuring the distribution router to send only a default route to the remote router, you must use the ip classless command on the remote router. By default, the ip classless command is enabled in all Cisco IOS images that support the EIGRP stub routing feature.

Without the stub feature, even after the routes that are sent from the distribution router to the remote router have been filtered or summarized, a problem might occur. If a route is lost somewhere in the corporate network, EIGRP could send a query to the distribution router, which in turn sends a query to the remote router even if routes are being summarized. If there is a problem communicating over the WAN link between the distribution router and the remote router, an EIGRP stuck in active (SIA) condition could occur and cause instability elsewhere in the network. The EIGRP stub routing feature allows a network administrator to prevent queries from being sent to the remote router.

Dual-Homed Remote Topology

In addition to a simple hub-and-spoke network where a remote router is connected to a single distribution router, the remote router can be dual-homed to two or more distribution routers. This configuration adds redundancy and introduces unique issues, and the stub feature helps to address some of these issues.

A dual-homed remote router has two or more distribution (hub) routers. However, the principles of stub routing are the same as they are with a hub-and-spoke topology. Figure 30-5 shows a common dual-homed remote topology with one remote router, but 100 or more routers could be connected on the same interfaces on distribution router 1 and distribution router 2. The remote router uses the best route to reach its destination. If distribution router 1 experiences a failure, the remote router can still use distribution router 2 to reach the corporate network.

Figure 30-5 Simple Dual-Homed Remote Topology

Figure 30-5 shows a simple dual-homed remote with one remote router and two distribution routers. Both distribution routers maintain routes to the corporate network and stub network 10.1.1.0/24.

Dual-homed routing can introduce instability into an EIGRP network. In Figure 30-6, distribution router 1 is directly connected to network 10.3.1.0/24. If summarization or filtering is applied on distribution router 1, the router advertises network 10.3.1.0/24 to all of its directly connected EIGRP neighbors (distribution router 2 and the remote router).

Figure 30-6 Dual-Homed Remote Topology With Distribution Router 1 Connected to Two Networks

Figure 30-6 shows a simple dual-homed remote router where distribution router 1 is connected to both network 10.3.1.0/24 and network 10.2.1.0/24.

If the 10.2.1.0/24 link between distribution router 1 and distribution router 2 has failed, the lowest cost path to network 10.3.1.0/24 from distribution router 2 is using the remote router (see Figure 30-7). This route is not desirable because the traffic that was previously traveling across the corporate network 10.2.1.0/24 is now sent across a much lower bandwidth connection. The over utilization of the lower bandwidth WAN connection can cause a number of problems that might affect the entire corporate network. The use of the lower bandwidth route that passes using the remote router might cause WAN EIGRP distribution routers to be dropped. Serial lines on distribution and remote routers could also be dropped, and EIGRP SIA errors on the distribution and core routers could occur.

Figure 30-7 Dual-Homed Remote Topology with a Failed Route to a Distribution Router

It is not desirable for traffic from distribution router 2 to travel through any remote router in order to reach network 10.3.1.0/24. If the links are sized to handle the load, it is acceptable to use one of the backup routes. However, most networks of this type have remote routers located at remote offices with relatively slow links. This problem can be prevented if proper summarization is configured on the distribution router and remote router.

It is typically undesirable for traffic from a distribution router to use a remote router as a transit path. A typical connection from a distribution router to a remote router has much less bandwidth than a connection at the network core. Attempting to use a remote router with a limited bandwidth connection as a transit path generally produces excessive congestion to the remote router. The EIGRP stub routing feature can prevent this problem by preventing the remote router from advertising core routes back to distribution routers. Routes learned by the remote router from distribution router 1 are not advertised to distribution router 2. Because the remote router does not advertise core routes to distribution router 2, the distribution router does not use the remote router as a transit for traffic destined for the network core.

The EIGRP stub routing feature can help to provide greater network stability. In the event of network instability, this feature prevents EIGRP queries from being sent over limited bandwidth links to nontransit routers. Instead, distribution routers to which the stub router is connected answer the query on behalf of the stub router. This feature greatly reduces the chance of further network instability due to congested or problematic WAN links. The EIGRP stub routing feature also simplifies the configuration and maintenance of hub-and-spoke networks. When stub routing is enabled in dual-homed remote configurations, it is no longer necessary to configure filtering on remote routers to prevent those remote routers from appearing as transit paths to the hub routers.

Caution EIGRP stub routing should only be used on stub routers. A stub router is defined as a router connected to the network core or distribution layer through which core transit traffic should not flow. A stub router should not have any EIGRP neighbors other than distribution routers. Ignoring this restriction causes undesirable behavior.

---

Note Multi-access interfaces, such as ATM, Ethernet, Frame Relay, ISDN PRI, and X.25, are supported by the EIGRP stub routing feature only when all routers on that interface, except the hub, are configured as stub routers.

EIGRP Stub Routing Configuration Tasks

To configure EIGRP stub routing, perform the tasks described in the following sections. The tasks in the first section are required; the task in the last section is optional.

• Configuring EIGRP Stub Routing (required)
• Verifying EIGRP Stub Routing (optional)

Configuring EIGRP Stub Routing

To configure a remote or spoke router for EIGRP stub routing, perform this task:

Verifying EIGRP Stub Routing

To verify that a remote router has been configured as a stub router with EIGRP, use the
show ip eigrp neighbor detail command from the distribution router in privileged EXEC mode. The last line of the output shows the stub status of the remote or spoke router. The following example shows output is from the show ip eigrp neighbor detail command:

Monitoring and Maintaining EIGRP

To delete neighbors from the neighbor table, use the following command:

To display various routing statistics, use the following commands:

EIGRP Configuration Examples

This section contains the following examples:

• Route Summarization Example
• Route Authentication Example
• Stub Routing Example

Route Summarization Example

The following example disables autosummarization and causes EIGRP to summarize network 10.0.0.0 out Ethernet interface 0 only:

Note You should not use the ip summary-address eigrp summarization command to generate the default route (0.0.0.0) from an interface because it creates an EIGRP summary default route to the null 0 interface with an administrative distance of 5. The low administrative distance of this default route can cause this route to displace default routes learned from other neighbors from the routing table. If the default route learned from the neighbors is displaced by the summary default route, or if the summary route is the only default route present, all traffic destined for the default route does not leave the router. Instead, this traffic is sent to the null 0 interface where it is dropped.

The recommended way to send only the default route out a given interface is to use a distribute-list command. You can configure this command to filter all outbound route advertisements sent out the interface with the exception of the default (0.0.0.0).

Route Authentication Example

The following example enables MD5 authentication on EIGRP packets in autonomous system 1:

Router A

Router B

Router A accepts and attempts to verify the MD5 digest of any EIGRP packet with a key equal to 1. It also accepts a packet with a key equal to 2. All other MD5 packets are dropped. Router A sends all EIGRP packets with key 2.

Router B accepts key 1 or key 2, and sends key 1. In this scenario, MD5 authenticates.

Stub Routing Example

A router that is configured as a stub with the eigrp stub command shares connected and summary routing information with all neighbor routers by default. Four optional keywords can be used with the eigrp stub command to modify this operation:

• receive-only
• connected
• static
• summary

This section provides configuration examples for all forms of the eigrp stub command. The eigrp stub command can be modified with several options, and these options can be used in any combination except for the receive-only keyword. The receive-only keyword restricts the router from sharing any of its routes with any other router in that EIGRP autonomous system, and the receive-only keyword does not permit any other option to be specified because it prevents any type of route from being sent. The three other optional keywords (connected, static, and summary) can be used in any combination but cannot be used with the receive-only keyword. If any of these three keywords is used individually with the eigrp stub command, connected and summary routes are not sent automatically.

The connected keyword permits the EIGRP stub routing feature to send connected routes. If the connected routes are not covered by a network statement, it may be necessary to redistribute connected routes with the redistribute connected command under the EIGRP process. This option is enabled by default.

The static keyword permits the EIGRP stub routing feature to send static routes. Without the static keyword, EIGRP does not send any static routes, including internal static routes that normally are automatically redistributed. It is still necessary to redistribute static routes with the redistribute static command.

The summary keyword permits the EIGRP stub routing feature to send summary routes. Summary routes can be created manually with the summary address command or automatically at a major network border router with the auto-summary command enabled. This option is enabled by default.

In the following example, the eigrp stub command is used to configure the router as a stub that advertises connected and summary routes:

In the following example, the eigrp stub connected static command is used to configure the router as a stub that advertises connected and static routes (sending summary routes is not permitted):

In the following example, the eigrp stub receive-only command is used to configure the router as a stub, and connected, summary, or static routes is not sent:

In the following example, the eigrp stub redistributed command is used to configure the router as a stub that advertises redistributed routes (sending connected, static, or summary routes is not permitted):

In the following example, the eigrp stub command is used to configure the router as a stub that advertises redistributed, static, connected and summary routes: