With the usage of Jira increasing, we notice that Jira is blocking out users for typically 120 to 300 seconds. The timing of this blocking seems to coincide with the garbage collector runs of the Java VM.
This is increasingly critical as users complain that when they have entered lengthy issue descriptions, they loose all their work (sometime in excess of half an hour). The number of user complaints is getting significant.
Is there a method to
a) determine the root cause of these short outages
b) fix it
Martin, Advantest
