1 of 1 people found this helpful
The cluster is not build correctly.
JGroups show that the incomming message from node2 is discarded because the current node1 does not have a cluster view which contains the node2
You should add a bit more information about your installation and maybe the relevant part of the logfile.
Please use the edit function for the comment and attach the files instead of posting
I have followed this guide, http://blog.akquinet.de/2012/06/21/clustering-in-jboss-as7eap-6/?blogsub=confirming#subscribe-blog.
Then tested by deploying sample,
This worked fine,
But when I deployed my application, I saw this in log and there is no other details.
I am using SLSBs and they are without @Clustered annotation. Will this cause issue?
I got following message in node 1 log:
INFO [org.infinispan.remoting.transport.jgroups.JGroupsTransport] (Incoming-7,shared=udp) ISPN000093: Received new, MERGED cluster view: MergeView::[node1/ejb|3] (2) [node1/ejb, node2/ejb], 2 subgroups: [node1/ejb|1] (1) [node2/ejb], [node1/ejb|2] (1) [node1/ejb]
How your application and configuration look like?
To start a simple cluster you can use the default standalone-ha.xml. Start "standalone.sh -c standalone-ha.xml".
I'm not sure when the bahaviour has changed, but @Clustered is deprecated and the application is deployed clustered if you enable HA configuration.
So maybe additional information will give a glue.
This isn't necessarily indicative of a problem. If you start 2 nodes simultaneously, it is possible that each member forms a separate cluster, which is subsequently merged. If this happens, and a message is broadcast to the group's multicast socket, it will be received and dropped, until the merge completes. This is perfectly normal.
I checked the log at that time and found it merged later as you mentioned. I checked by starting second node after starting first one and there is no warning showing now and both working fine.
I referred this http://blog.akquinet.de/2012/06/21/clustering-in-jboss-as7eap-6/?blogsub=confirming#subscribe-blog and found @Clustered used for EJBs.
I am using HA profile, so we don't need @Clustered and all beans will be clustered by default, right?