To validate polling and thresholds, do the following using the Polling and Thresholds Console:

  • Cause a failure by physically removing a cable in a discovered portion of the network. Check that the correct notification is received at the Global Console and that the topology map indicates the failure. Note that you might have to wait a few polling cycles to see the correct root-cause analysis.

  • Reduce threshold settings to very low levels or zero. Typically, reducing port or interface performance settings works well for validating Performance Manager deployments. For IP Availability Manager deployments, reduce the RestartTrapThreshold connectivity setting for a router or switch and cause a warmstart on the corresponding type of system. Obviously, choose equipment and a time frame when you will not interrupt your network users. Verify that the correct notification is received at the Global Console indicating that the threshold was exceeded.

    Validating that the levels that you chose for the thresholds are appropriate for your deployment is much more difficult. Start by using the defaults and adjust them upward or downward as you gain experience with the equipment. When failures occur, particularly failures with little or no warning, determine if the failures were preceded by symptoms that could have been detected by lower threshold values. Then adjust the thresholds appropriately. This evaluation should be performed after all failures.