Tim, Thanks for the quick replies. I will try the fix from TRUNK on Monday.
In the meantime here are some further observations from my last tests with GA: I suppose part of the memory issues stem from the use of HSQL (for easier reproducability I use the out-of-the-box messaging configuration for the tests). It happens if one of the test subscribers has a very long session restart cycle time and the publishers produce messages faster than they can be consumed. As is to be expected memory usage starts to grow after more than 200.000 messages build up in the subscription. What I also observed when messages started to be persisted is that the JMS system locked up temporarily several times for about 6 to 8 minutes! After that time message publishing and consuming resumed. I suppose some timer task has performed some cleanup operation at that time, although I have no idea what it might be. View the original post : http://www.jboss.com/index.html?module=bb&op=viewtopic&p=4052154#4052154 Reply to the post : http://www.jboss.com/index.html?module=bb&op=posting&mode=reply&p=4052154 _______________________________________________ jboss-user mailing list jboss-user@lists.jboss.org https://lists.jboss.org/mailman/listinfo/jboss-user
