[ http://jira.jboss.com/jira/browse/JBREM-560?page=all ] Ron Sigal closed JBREM-560. --------------------------- Resolution: Done Recent changes to allow MultiplexServerInvoker to shut down upon experiencing unrecoverable i/o errors caused this problem. The acceptThread caught an error and called stop() before main thread called Connector.stop(). The result was that the socket group was shut down asynchronously in the acceptThread, causing a timing problem. Right after call to Connector.stop(), acceptThread hadn't finished shutting down socket group, causing the test for the absence of the socket group to fail. Changed MultiplexServerInvoker in two ways: 1. When a master invoker creates a virtual invoker, it sets a variable "hasMaster" to true on virtual invoker. 2. When run() encounters a non-recoverable error, it returns, but only calls stop() if hasMaster == true. Otherwise, let another thread call Connector.stop(). -- This message is automatically generated by JIRA. - If you think it was sent incorrectly contact one of the administrators: http://jira.jboss.com/jira/secure/Administrators.jspa - For more information on JIRA, see: http://www.atlassian.com/software/jira