@Harish Tuccapuram, what version of Jira are you running? I am having a similar thing go on with mine (7.11.2). The one difference I have is that I have my pool-max-size at 60. The symptoms I'll see are that BOTH DB connections and the CPU will go bonkers at the same time. Check to see if you're getting the CPU load high at the same time.
I've been mucking about with it to try to ameliorate the problem. I am running this containerized on AWS ECS. What I've done recently to try to get past this is add more CPU credits to the Jira container to give it more headroom.
I noticed that the baseline CPU usage dropped and I don't get the DB connection spikes (neither of which would recover if it went crazy. I still see some blips above the baseline 20 and occasional CPU spikes when something heavy happens but it's not gone off the deep end for me since that change (yet). Perhaps a similar change will help you? At the very least I'd give it more DB connections for more headroom there. I wonder if Jira doesn't get in a load tizzy where more is coming in than it can handle and can't get past the incoming when not enough room for outgoing.
My next test will be Monday when sprint planning is going on and the system is being pounded.
You must be a registered user to add a comment. If you've already registered, sign in. Otherwise, register and sign in.