[jboss-jira] [JBoss JIRA] Resolved: (JBCLUSTER-93) Clustered MBean disappears from cluster when stopping one cluster node

[Brian Stansberry (JIRA)]

     [ http://jira.jboss.com/jira/browse/JBCLUSTER-93?page=all ]

Brian Stansberry resolved JBCLUSTER-93.
---------------------------------------

    Resolution: Duplicate Issue

It's not the same as JBAS-3023, but I've created JBAS-3194 to track it in the AS project.  Am closing this issue; the bug should be tracked via JBAS-3194.

> Clustered MBean disappears from cluster when stopping one cluster node
> ----------------------------------------------------------------------
>
>                 Key: JBCLUSTER-93
>                 URL: http://jira.jboss.com/jira/browse/JBCLUSTER-93
>             Project: JBoss Clustering
>          Issue Type: Bug
>      Security Level: Public(Everyone can see) 
>         Environment: JBoss 4.0.3 ,  Cluster nodes run Kubuntu Linux 2.6.12-10-686,  192.168.0.X IP addresses
>            Reporter: Hannes Koller
>         Assigned To: Brian Stansberry
>             Fix For: Bugs
>
>
> Setup: 
> --------
> * Clustered MBean  (extends HAServiceMbeanSupport)
> * ProxyFactoryHA is used to create the proxies 
> (see refrenced forum thread - second page for detailed description of the MBean and associated jboss-service.xml)
> * Clustered MBean is deployed on several nodes (deployed manually, farming disabled)
> Behavior: 
> -------------
> Initially everything works ( Load Balancing works - invocations are distributed among the nodes ).  Bug occurs when one node undeploys it's replicant of the clustered MBean. All Instances of the MBean  disappear from the DistributedReplicantManager, when one node shuts down the MBean. 
> Core Reason: 
> ----------------------
> An MBean.STOPPING event is created when the MBean is undeployed. This event is propagated to EVERY replicant of the MBean (not just the instance on the node which is actually shut down) --> the containerIsAboutToStop()  method of the ProxyFactoryHA is called on every node, thus wrongly unregistering all MBean replicants at the DistributedReplicantManager. 

-- 
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