I will appreciate any insight regarding the behavior observed with JBoss 4.2.3 GA, JBoss Messaging and Worker Threads. The JBoss 4.2.3 GA is running on Windows XP (SP 3) and uses jdk1.6.0_12. The server is configured to use JBoss Messaging jboss-messaging-1.4.2.GA-SP1 (instead of JBoss MQ provided with JBoss 4.2.3 GA). Two mbean sections in jboss-service.xml were modified | to configure limited number of threads in the pool | 1. org.jboss.invocation.pooled.server.PooledInvoker | MaxPoolSize=15 | ClientMaxPoolSize=15 | SocketTimeout=60 | | 2. org.jboss.remoting.transport.Connector (Configuration subsection) | maxPoolSize=15 | clientMaxPoolSize=15 | socketTimeout=60 The Topic example supplied with the jboss-messaging-1.4.2.GA-SP1 examples directory was modified to simulate a client that is doing some work. A call to Thread.sleep() for 120 seconds was added just before releasing resources (i.e. before closing the InitialContext and connection in the finally block). Threads were monitored using jvisualvm (provided by the JDK). After starting the JBoss server there were 57 live threads, 35 daemon threads, there were no Worker Threads. To simulate multiple clients publishing to a Topic and listening for replies; 25 instances of the Topic example were started in quick succession. The live thread count on the JBoss server went up to 136 of which 49 were Worker Threads (in running state). After all the clients completed their sleep time (120 seconds) the client JVMs exited. At this point the JBoss server showed 106 live threads, 35 daemon threads. There were 49 Worker Threads in waiting state. All Worker threads remained in memory (in waiting state) even after an hour had passed with no activity (no requests) on the server. Is it ok to expect that some Worker threads are closed by the server bringing down the live threads count due to inactivity for an hour? Are there any parameters in JBoss that need to be changed to maintain the desired number of Worker Threads after a period of inactivity? Thanks for reading this post, any inputs or suggestions throwing light on the observed behavior is welcome. View the original post : http://www.jboss.org/index.html?module=bb&op=viewtopic&p=4263339#... Reply to the post : http://www.jboss.org/index.html?module=bb&op=posting&mode=reply&a...