[jboss-jira] [JBoss JIRA] (WFLY-3711) Topology updates of EJBClient ClusterContexts not being processed correctly after failover

[Richard Achmatowicz (JIRA)]

    [ https://issues.jboss.org/browse/WFLY-3711?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12993683#comment-12993683 ] 

Richard Achmatowicz commented on WFLY-3711:
-------------------------------------------

Another problem is that when topology updates arrive, we do not differentiate between these two types of topology updates:
- an update received from an already connected node which tells us that a new node (or nodes) have been added to the existing cluster topology
- an update from a newly connected node which represents a full cluster topology )(i.e. the contents of that nodes Registry entry for the cluster in question) 

At the moment, both of these update messages are handled by the same handler (ClusterTopologyUpdateHandler) which interprets the set of nodes passed as nodes which have been added to the cluster: the response is to add the nodes to the existing cluster context. This is different from interpreting the set of nodes passed as the nodes which make up the current state of the cluster: the correct response here is to make sure that the cluster context matches this set of nodes (and may require removing nodes which are no longer members. 

What should happen is that the two message types should be differentiated and handled differently (in the way described above).



> Topology updates of EJBClient ClusterContexts not being processed correctly after failover
> ------------------------------------------------------------------------------------------
>
>                 Key: WFLY-3711
>                 URL: https://issues.jboss.org/browse/WFLY-3711
>             Project: WildFly
>          Issue Type: Bug
>      Security Level: Public(Everyone can see) 
>          Components: Clustering
>    Affects Versions: 9.0.0.Beta1
>            Reporter: Richard Achmatowicz
>            Assignee: Richard Achmatowicz
>
> ClusterContexts are used by EJBClient to keep track of the current set of nodes in a cluster, so that if an EJBClient invocation fails on one node, it may failover to another node in the same cluster. The ClusterContext is made up of ClusterNodeManagers which are responsible for setting up the connections between the EJBClient and the nodes in the cluster.
> Cluster topology updates are sent to registered EJBClients whenever the cluster topology changes (nodes join, nodes leave a cluster). Thse topology updates are processed on the client side by ClusterTopologyUpdateHandler and are used to update the current contents of the associated ClusterContext held on the client.
> The current implementation of the handling of topology updates does not correctly handle the addition/removal of ClusterNodeManagers from the cluster context - namely, rather than check whether or not a new ClusterNodeManager really needs to be added, ClusterNodeManagers are added for every node in the received topology update, leading to many unnecessary EJBReceivers and their channels being created.
> The logs show that up to 18 cluster node manager instances may be created, have their EJBReceivers registered and channels to the remote node created, when only one node has been added to the cluster.



--
This message was sent by Atlassian JIRA
(v6.2.6#6264)