Worst case is the highest utilization that a particular interface experiences over all the failure sets and traffic levels that you selected. Cisco Crosswork Planning determines which combination of failures would cause this worst-case utilization.

The default analysis is to identify up to 10 failures for the worst-case utilization on each interface in the network. Alternatively, you can record failures causing utilizations within a specified percent of the worst-case utilization. To control the number of threads that Cisco Crosswork Planning processes in parallel when examining failure scenarios, set the Maximum number of threads field. For details on how to run the Simulation analysis, see Run Simulation Analysis.

Upon finishing the analysis, Cisco Crosswork Planning switches to the Worst-Case traffic view and updates the plot to simultaneously display the worst-case utilization for all interfaces.

In addition to Plot view changes, the following columns are also updated in the Interfaces and Circuits tables upon finishing the Simulation analysis:

• WC util—The worst-case utilization for that interface. The worst-case for a circuit is defined to be whichever of the worst cases of the two constituent interfaces results in the larger utilization. Thus, for circuits, this value is the larger of the WC util values for the two interfaces in the circuit.

• WC traffic—The actual traffic (Mbps) through the interface under the worst-case scenario.

• WC traff level—The traffic level under which this worst-case scenario occurs.

• WC failure—List of one or more failures that cause the worst-case failure of the circuit. An easier way to read this list is to select an interface and use > Fail to WC.

  If you record failures causing utilizations within a given percent of worst case, this column shows QoS violations as a percent. (See Identify Worst-Case QoS Violations.) If the number is positive, then the allotted capacity has been surpassed. If negative, the capacity has not been surpassed. For example, if a circuit has 10,000 Mbps capacity, and if the amount of traffic on it as a result of three different failures is 11,000, 8000, and 4000 Mbps, the utilizations are 10%, –20%, and –60%, respectively, and in descending order.

  

• WC service class—The service class for which this worst-case scenario occurs.

For information on running a Simulation analysis with QoS, see Identify Worst-Case QoS Violations.

For information on worst-case calculations for VPNs, see Simulate VPN.

When running Simulation analysis (see Run Simulation Analysis), you have the option to simulate worst-case latency for each demand in the plan. Cisco Crosswork Planning calculates the maximum latency of each demand under the failure scenarios selected. The result does not depend on service classes or traffic levels because demand routing is independent of these plans. The simulation also records the failures that cause this maximum latency.

The following columns in the Demands table are updated when you check the Calculate demand worst-case latency check box (under the "Calculate worst-case utilization interface" section) while running the Simulation analysis tool:

• WC latency—The highest demand latency over all failure scenarios in the analysis.

• WC latency failures—The failures that caused this worst-case latency. Up to 10 failures are identified.

Cisco Crosswork Planning captures the latencies of each demand for each failure case included in Simulation analysis using the following two options:

• Record failures causing demand latency within __ % of worst case—Records failures causing demand latency within the specified percentage range of the worst-case latency. Default is 0. If you enter 0, only the worst case latency failures are recorded.

• Record up to __ failure scenarios on demand latency—Maximum number of failure scenarios to record per demand. Default is 1.

If you record failures causing demand latency within a given percent of worst case, the WC latency failures column shows the WC latency along with failure scenarios.

Each time Simulation analysis is run, a report is automatically generated. You can access this information at any time by choosing Actions > Reports > Generated reports and then clicking the Simulation Analysis link in the right panel. Note that new reports replace the previous ones.

The Options tab displays the input parameters used for the simulation analysis.

The Summary tab details the options used in the analysis and summarizes the most important problems identified, such as QoS violations and latency bound violations.

The Max Utilization tab shows the impact of failures on maximum utilization in the form of a pie chart.

The Simulations table lists each simulation that was performed in the Simulation analysis (Table 1).

The Failure impact view is available upon running a Simulation analysis (Example Failure Impact). The plot in this view colors the nodes and circuits according to the maximum utilization level that would be caused elsewhere in the network should the node or circuit fail. The color indicates the resulting utilization and severity of the congestion.

Example: In the Failure impact view, a sjc-lax circuit has a utilization of 90-100% and its color representation is orange red. This means that if sjc-lax were to fail, one or more interfaces would react by exceeding a 90% utilization level and correspondingly, would turn orange red in the plot.

The Node, Interface, and Circuit tables contain the Failure impact and Failure impact interface columns. In the Interfaces table, the information describes the failure impact of the circuit containing the interface.

• Failure impact—The failure impact of each node or circuit. For example, if the value is 80%, it means that if this node or circuit failed, the resulting traffic utilization on one or more interfaces would exceed 80%.

• Failure impact interface—The interface that will experience the highest utilization as a result of the node or circuit going down.

Format = if{Node|Interface}

Example: if{cr2.sjc|to_cr1.kcy} means if the circuit goes down, it will have the greatest traffic impact on the cr2.sjc to cr1.kcy interface.

In the network plot, the Site borders show the maximum utilization level that would be caused elsewhere in the network should nodes within it fail or should intra site circuits within it fail.

Note	The Failure impact view only shows the impact of circuit and node failures. It does not show failures of other objects.

Simulation Analysis tool allows parallelizing the computaion and hence arrive at a faster result for large network models.

Example: If there are 10000 circuits in a network model, and there are 10 different engines available, the tool can be used to break up the network model into 10 partitions with each partition handling 1000 failure scenarios. This results in 10 different result files. The results from each of these independent runs are merged together to obtain the final result.

Execute parallelization:

Use the CLI tool to execute parallelization. For details, see Run Tools or Initializers Using CLI.

1. From the main menu, choose Job Manager.

2. Click > Using CLI.

3. Choose Simulation analysis and click Next.

4. Select the network model in which you want to run the simulation analysis and click Next.

5. In the "Simulation Analysis input options" field, use the following:

   -failure-sets <failure-sets> -num-partitions <number-of-partitions> -num-threads <number-of-threads> -partition-index <partition-index> -result-file <result-filename>

   where

   • -num-partitions– Number of partitions of the failure scenarios. Each partition has an associated set of failure scenarios and is identified by an index ranging from 0 up to the number of partitions minus 1. Default is 1.

   • -partition-index– Simulate the set of failure scenarios belonging to the specified partition. Default is 0.

   • -result-file– If specified, the simulation analysis report results are written to this file. Can be *.txt or *.db file.

For more information on running tools and initializers using CLI, see Run Tools or Initializers Using CLI.

import os
import sys

cmd = "merge_sim_analysis -plan-file {0} -partial-results {1} -out-file out-plan.txt ".format( sys.argv[1], sys.argv[2])
print(cmd)
os.system(cmd)

run_cli_merge.py input_planfile.pln res_0.txt,res_1.txt,res_2.txt