9 Replies Latest reply on Jul 24, 2015 7:44 AM by Yanick Girouard

    What is the best and most optimized way to monitor and log usage of the multiple app sever threads?

    Yanick Girouard

      I'm looking for a way to monitor and log the statistics of work item, nsh proxy threads and db connections (basically everything shown in the Memory Monitor probes) almost (if not) in real time.


      The current Memory Monitor dumped to the app server logs are too spaced out to be really useful, and parsing the logs to extract those in real time can be a little tedious (and costly in terms of performance).


      To put you in context, we have 4 app severs, and we're still trying to tweak the different capacity and thread related settings. I suspect we're getting dangerously close to attaining the limit of some of the threads, but because the different memory monitor dumps in each appserver's logs are too spaced out, and also not dumped simultaneously, it's hard to find concurrency issues where all threads are maxed out across all app servers. Sometimes the concurrency issues may last only a few seconds, but it's still enough to make some jobs fail.


      I would need a way to be able to get close-to-realtime statistics that I could easily query later (i.e. stored in a db somewhere).


      What would be the best way and tool to do this? Keep in mind that whatever the solution is, it can't affect the performance of the app servers.