[jboss-jira] [JBoss JIRA] Commented: (JBMESSAGING-1890) Recovery configuration in a cluster is too burdensome

[Justin Bertram (JIRA)]

    [ https://issues.jboss.org/browse/JBMESSAGING-1890?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12615284#comment-12615284 ] 

Justin Bertram commented on JBMESSAGING-1890:
---------------------------------------------

I opened JBMESSAGING-1731 awhile back to make it possible for users who wanted to write their own cluster-aware recovery module.  However, this functionality already exists in EAP 5 via the JGroups channel factory.  The 2 JGroups channels used by JBoss Messaging are exposed via JMX (e.g. jboss.jgroups:cluster=DefaultPartition-JMS-CTRL,type=channel and jboss.jgroups:cluster=DefaultPartition-JMS-DATA,type=channel).  Therefore a possible solution to this problem in EAP 5+ is to write a new cluster-aware recovery module using this information.

Another possible solution would be to simply *not* fail-over in-doubt XA transactions to other nodes in the cluster.

> Recovery configuration in a cluster is too burdensome
> -----------------------------------------------------
>
>                 Key: JBMESSAGING-1890
>                 URL: https://issues.jboss.org/browse/JBMESSAGING-1890
>             Project: JBoss Messaging
>          Issue Type: Task
>          Components: JMS Clustering
>    Affects Versions: 1.4.8.SP1
>            Reporter: Justin Bertram
>            Assignee: Yong Hao Gao
>
> The recovery configuration describe on http://community.jboss.org/wiki/Non-serializableXAResourceErrorFromJBossMessagingJBM-ClusteredSetup is too burdensome, especially for large clusters.  When a new node is added to the cluster then *every* node in the cluster needs a configuration change.

--
This message is automatically generated by JIRA.
For more information on JIRA, see: http://www.atlassian.com/software/jira