[
https://issues.apache.org/activemq/browse/AMQ-1283?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=41710#action_41710
]
Trevor Pounds commented on AMQ-1283:
------------------------------------
Here is another post that may be related. It seems that the fix suggested by
Pete Schwamb fixed his issue. The fix was checked into the SVN repo r636078.
Thread
http://www.nabble.com/5.0-%28and-later%29-Queue-stops-dispatching-td15987081s2354.html#a15987081
Fisheye
http://fisheye6.cenqua.com/browse/activemq/trunk/activemq-core/src/main/java/org/apache/activemq/kaha/impl/index/DiskIndexLinkedList.java?r1=598330&r2=636078
> Messages aren't being delivered in a clustered environment
> ----------------------------------------------------------
>
> Key: AMQ-1283
> URL: https://issues.apache.org/activemq/browse/AMQ-1283
> Project: ActiveMQ
> Issue Type: Bug
> Components: Broker
> Affects Versions: 4.1.1
> Environment: Java 1.5.0_04, Windows XP
> Reporter: Matthew Vincent
> Fix For: 5.2.0
>
>
> Messages aren't always delivered in a clustered environment.
> If you have a cluster of three ActiveMQ nodes, and you send the cluster a
> constant stream of messages, and randomly restart individual ActiveMQ nodes,
> you can enter a state whereby an ActiveMQ node has undelivered messages
> sitting in the message store. These messages will not be delivered, unless
> that ActiveMQ node is restarted.
> This appears to be a consequence of the fact that a Queue immediately
> dispatches a message received to any available subscribers, but ignores any
> exceptions that result from the actual send.
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
