2 Replies Latest reply on Dec 14, 2012 8:05 PM by silverdev Branched to a new discussion.

    Session replication using AS7 clustering, S3 PING, Amazon EC2

    silverdev

      Hello all,

       

      I am trying to setup session replication using AS7 clustering, S3 PING and Amazon EC2 for the Seam booking EAR app.

       

      So far, I have managed to get session replication working with two different local servers running on the same network using S3 ping and AS7:

      1. The S3 bucket is created with two files.
      2. I can restart one server and continue my browser session.

      I am running into a problem when trying to get this same setup running on Amazon EC2. Here's how far I could get on the Amazon cloud:

      1. I start JBoss using the following command:standalone.sh -c standalone-ha.xml -Djboss.server.base.dir=standalone -b 10.X.Y.Z -Djboss.default.jgroups.stack=tcp -Djboss.node.name=node1 &
      2. The S3 bucket is created with two files. Relevant part of standalone-ha.xml:
        <protocol type="S3_PING">                    
        <property name="access_key">                        
        xxx                    
        </property>                    
        <property name="secret_access_key">                        
        xxx                    
        </property>                    
        <property name="prefix">                        
        jbossas7-s3ping                    
        </property>                    
        <property name="timeout">                        
        60000                    
        </property>                
        </protocol>
        
      3. JGroups logs says:
        09:59:23,254 INFO  [org.jboss.as.clustering.jgroups] (ServerService Thread Pool -- 37) JBAS010260: Activating JGroups subsystem.
        10:00:08,631 INFO  [org.infinispan.remoting.transport.jgroups.JGroupsTransport] (pool-15-thread-1) ISPN000078: Starting JGroups Channel
        10:00:08,636 DEBUG [org.infinispan.remoting.transport.jgroups.JGroupsTransport] (pool-15-thread-1) New view accepted: [node1/web|0] [node1/web]
        10:00:08,636 INFO  [org.infinispan.remoting.transport.jgroups.JGroupsTransport] (pool-15-thread-1) ISPN000094: Received new cluster view: [node1/web|0] [node1/web]
        10:00:08,637 INFO  [org.infinispan.remoting.transport.jgroups.JGroupsTransport] (pool-15-thread-1) ISPN000079: Cache local address is node1/web, physical addresses are [10.194.93.149:7600]
        10:00:08,638 DEBUG [org.infinispan.remoting.transport.jgroups.JGroupsTransport] (pool-15-thread-1) Waiting on view being accepted
        

      Note:

      1. I have opened ICMP,TCP ports for EC2 Security Groups on both EC2 instances
      2. I have disabled firewall on both instances
      3. I can telnet from instance to the other on port 7600.
      4. I have attached my AS7 config.

       

      It seems like the two EC2 instances do not talk to each other. I can't find the following message in the EC2 logs:

      [org.jgroups.protocols.FD_SOCK] (Incoming-1,null) VIEW_CHANGE received: [node1/web, node2/web]
      

       

      Can anyone help?