I'm actually not sure why ActiveMQThreadFactory needs a thread group at all. As long as the threads are named appropriately I think there doesn't need to be a group and that would solve this leak. Otherwise we'd have to figure out a way to call destroy() on the group when pool is shutdown and that isn't obvious.
Justin ----- Original Message ----- From: "Denis V. Kirpichenkov" <[email protected]> To: [email protected] Sent: Thursday, May 4, 2017 12:48:43 PM Subject: ARTEMIS-874 ThreadGroup memory leak Hello, All! Recently I faced with a memleak described in https://issues.apache.org/jira/browse/ARTEMIS-874. And my question is why does ActiveMQThreadFactory.defaultThreadFactory() create its own ThreadGroup? May be it is fine to work without it?
