Github user franz1981 commented on the issue:
https://github.com/apache/activemq-artemis/pull/2481
I will be able to look into that in the next days :)
I can already ask you to collect time to safepoints/GC pauses if possible
ie -XX:+PrintGCApplicationStoppedTime.
2 minutes seems a too long period TBH, but worth taking a look if you rely
on MAPPED journal and/or paging a lot, given that major page faulting can
causes long stall similar to very long full GC, but TBH nothing so long (2
minutes is a lot!).
As an additional suggestion you could run a Java program that just use one
core and jmeasure the elapsed time between 2 consecutive nanoTime calls,
recording in which wall-clock time a back-ward "drift" has happened to check if
a broker shutdown has happened near the same time, makes sense?
---
