As a followup, a few observations: The exception stack traces show that org.jboss.invocation.jrmp.server.JRMPInvoker_Stub.invoke() is used, while I believe JRMPHAInvoker should have been used? Also, I replaced SingleRetryInvoker with RetryInvoker. Shouldn't my client wait for a connection indefinately instead of issuing a connection refused? I am starting to suspect that my clustered container isn't used at all. Is there a way to check this? View the original post : http://www.jboss.com/index.html?module=bb&op=viewtopic&p=3983420#... Reply to the post : http://www.jboss.com/index.html?module=bb&op=posting&mode=reply&a...