This issue occurred with JBoss 5.1.0 GA.

It appears that the jboss session replication has resulted in all of the https threads being in WAITING or BLOCKED state due to the jboss https replication (see attached thread dump). At the time that this issue was detected there were no active users on the system.

I've also attached the jgroups-channelfactory-stacks.xml file that is being used.  The problem occurred while using the TCP stack with TCPPING and jboss shunning disabled.  The cluster has 2 members (including the local host).

Is there a patch or anything in the jboss configuration that can be changed to prevent this issue from occurring?