I've done another run of the testsuite with smaller pool sizes JRA and Remoting and thread reaping turned on for Remoting. I've attached some comments and a new report file to the original JIRA issue (sorry for the back and forth). The reduced pool sizes did not affect the passing/failing of tests as far as I can tell, and yet reduced the threading resources considerably - active thread count down by about 1/3. I've gone through and added the (+) as before and confirmed that WorkerThreads and WorkManager threads are being reclaimed. This sort of change alleviates the (original) problem I was having running the testsuite on HPUX where I was getting OOME: cannot allocate native thread due to no memory left for allocating thread stacks. In that case, the problem was also related to a very high -Xmx setting which left less memory available outside the JVM for allocating thread stacks. Lowering the -Xmx value on the 'all' configuration from 512m to 384m (which is closer to the AS 4.2/4.3 value of 256m) allowed more threads to be created and virtually eliminated the problem. View the original post : http://www.jboss.org/index.html?module=bb&op=viewtopic&p=4222136#... Reply to the post : http://www.jboss.org/index.html?module=bb&op=posting&mode=reply&a...