During testing cluster failover, I discovered that I often got messages like this in the log file of the broker which the new broker contacts to catch up: -------------------------------------------------------------------------------- 2009-mar-10 08:41:16 warning Session was not closed cleanly 2009-mar-10 08:41:16 warning Ignoring frame while closing connection: Frame[BEbe; channel=71; {SessionTimeoutBody: timeout=0; }] 2009-mar-10 08:41:16 warning Session was not closed cleanly 2009-mar-10 08:41:16 warning Session was not closed cleanly 2009-mar-10 08:41:16 warning Ignoring frame while closing connection: Frame[BEbe; channel=70; {SessionCommandPointBody: command-id=0; command-offset=0; }] 2009-mar-10 08:41:16 warning Ignoring frame while closing connection: Frame[BEbe; channel=70; {SessionTimeoutBody: timeout=0; }] 2009-mar-10 08:41:16 warning Session was not closed cleanly 2009-mar-10 08:41:16 warning Ignoring frame while closing connection: Frame[BEbe; channel=71; {SessionTimeoutBody: timeout=0; }] 2009-mar-10 08:41:17 warning Session was not closed cleanly 2009-mar-10 08:41:17 warning Session was not closed cleanly 2009-mar-10 08:41:17 warning Ignoring frame while closing connection: Frame[BEbe; channel=71; {SessionCommandPointBody: command-id=0; command-offset=0; }] 2009-mar-10 08:41:17 warning Ignoring frame while closing connection: Frame[BEbe; channel=71; {SessionTimeoutBody: timeout=0; }] 2009-mar-10 08:41:17 warning Session was not closed cleanly -------------------------------------------------------------------------------- The newly attached broker will just give a "(READY) caught up, active cluster member" log message when this error appears in the other broker. It looks like something is wrong, but according to Alan, this is expected. To reproduce it, setup openais and start a broker with --cluster-name. Then start a client which implements FailoverManager against the broker. Add new brokers to the same cluster. If these warnings really are harmless, they should not be logged.
May be related to bug 493016
Fixed all warnings during the update process in r927383 r927384 r927385
Technical note added. If any revisions are required, please edit the "Technical Notes" field accordingly. All revisions will be proofread by the Engineering Content Services team. New Contents: When a broker joins a cluster, it no longer writes unnecessary messages to the log file.
An advisory has been issued which should help the problem described in this bug report. This report is therefore being closed with a resolution of ERRATA. For more information on therefore solution and/or where to find the updated files, please follow the link below. You may reopen this bug report if the solution does not work for you. http://rhn.redhat.com/errata/RHSA-2010-0773.html