Display the SST Metric Data Correctly

Please refer to TS016456569. When the app system down happened due to max heap usage, the SST metrics display the heap or Default executor threads reached the max by searching a small window (30 minutes) under application server JVM metrics. However, if the search display window increases to 3 days, SST cannot display such critical events. The support team mentioned this is due to the metrics will be averaged out with the longer search window. Refer to the below comments.

What we are expecting for from SST metric is to search past 3 days or even longer time, and see whether the system was ever down and reached max threads or heap. And the current SST cannot honor that. This will make the production and preprod SST monitoring useless for displaying these critical events.

To provide an update, I tried this from my end and was able to see the same behaviour you reported. From checking with my team on the issue reported on this case, this seems to be expected behaviour.

For the Default executor threads dashboard for Application server JVM metrics on SST, which displays the app server's thread count within a specific timeframe, it seems the data is aggregated based on the time frame that is selected and then averaged out. So if you choose a longer time frame, you will see the average of the thread count data, rather than the exact thread count information.

This would help to explain why you are seeing your data showing up in this way when selecting a wider timeframe. As, from my understanding, this behaviour is expected within SST, you can log an enhancement request if you would like this behaviour to be changed for a future release. Our SST team will then process this request and can possibly look into adding as an enhancement within a future release. I have added the link for doing so below for your reference.