Problem Definition

In JIRA Datacenter cache for Users and Group is a core for many functionality. Cache flush for them is vert expensive and lead to significant performance degradation of the whole cluster.

Suggested Solution

Add INFO logging event for each cache flush event and additional DEBUG logging with stackrace to track source the flush event.

Implemented solution

• In fixed JIRA version, there is a logging for cache flush events.

  2017-01-25 23:55:19,583 http-nio-8090-exec-1 INFO      [c.a.c.e.c.a.j.c.e.ofbiz.EagerOfBizUserCache.userCache] Cache com.atlassian.jira.crowd.embedded.ofbiz.EagerOfBizUserCache.userCache was flushed
  

• Please note that this logging has no performance impact.
• How to enable logging for cache flush events:
  1. Navigate to Cog Icon > System > Logging and profiling
  2. Find com.atlassian.cache.event and set it to INFO level