Verify results

If AppMon is already running in a staging environment or as a POC, be sure to double-check your results and compare them with your requirements. To help you with this, AppMon provides several system performance metrics that you can use to verify your deployment. These metrics, which you can display in specialized AppMon self-monitoring dashboards, are listed and described in the table below. See Deployment Health Dashboards for more information about self-monitoring dashboards.

Deployment Verification Metrics

Metric Comments
Skipped Events per Interval Must be 0 or you receive incomplete PurePaths. If this happens, check the MPS metric, the send queue buffer size of your Collectors, and the Bytes Received per Interval metric.
MPS (Measures or Nodes per Second) Compare the values at Agent/Collector and Server level with your estimate. If they are much higher than expected, the instrumentation rules should be justified or you underestimated the transaction load. You can use the Completed PurePaths per Interval or the PurePath Length measures to decide what is causing the problem.
Bytes Received per Interval Compare these values at the Collector level with the physical bandwidth available to make sure that this is not a bottleneck.
Send Buffer Queue Used/Committed Size Check whether the used size consistently falls below the committed size. Whenever the used buffer reaches the committed size, skipped events occur. In some instances, the solution is as simple as increasing send queue buffer size.
PurePath Buffer Duration Check this value to determine the retention period of PurePath analysis data. This value is computed as the difference between the start time of the oldest PurePath and the end time of the latest PurePath in memory. As a result, it increases after the Server starts and stabilizes after the Server's buffers are filled.