Github user franz1981 commented on the issue:
https://github.com/apache/activemq-artemis/pull/1800
@gaohoward
I understand, but often a failed test is expected to fail cleanly (without
having leaking threads): having that information ignored could affect anyway
the other tests and probably it is hiding leaks that is better to know (eg
oracle JDBC has leaking driver clean threads that is better to have it reported
on failed tests too: thay won't stop never even if the thread is killed
manually).
Have I misunderstood the change?
---
