1 2 3 Previous Next 35 Replies Latest reply on Aug 25, 2012 10:06 AM by ranga033 Go to original post
      • 30. Re: mod_cluster 1.2 + JBossAS 7.1 problem
        johanvermeij

        It could be related to this because I don't really know what I am doing here.

         

        I run the master and slaves as follows:

         

        $JBOSS_HOME/bin/domain.sh -Djgroups.bind-addr=192.168.2.3

         

        (I read somewhere that you should bind jgroups to the ext_ip_address. After I added the -D warning displays from jgroups about dropped messages stopped appearing)

         

        -------------------------

         

        I just stopped the HP server and now throughput is super fast using the PC and the iMac

        • 31. Re: mod_cluster 1.2 + JBossAS 7.1 problem
          johanvermeij

          I thought it was the database on the HP so I moved that to a PC under XP

          The HP is still acting up though

          As soon as I start the JBoss instance errors pop up and processing grinds to a halt

          And when I shut down the instance performance is back to normal

           

          It is incredible to see 4 different machines under 3 different  O/S working like this

           

          JBoss never ceases to amaze me

          • 32. Re: mod_cluster 1.2 + JBossAS 7.1 problem
            ranga033

            Why do you explicitly bind the JGroups address . Does it not work without that ? I dont' think this is required.  JGroups will auto-detect the nodes and form the cluster.

             

            Also, I would suggest you create a new thread under AS7 discussion(currently it is under mod_cluster). You would find more experts in each of the modules there and get specific information.

             

            One more thing- are you using the same JGroups configuration for all the machines ?

            • 33. Re: mod_cluster 1.2 + JBossAS 7.1 problem
              johanvermeij

              Stupid had the firewall on on the HP server

               

              All ok now after I turn it off

               

              I have removed the explicit bind for the JGroups address

               

              Thanks and apologies for wasting your time

              • 34. Re: mod_cluster 1.2 + JBossAS 7.1 problem
                johanvermeij

                Early this morning I was ready for some serious performance testing and I fell into a hole again.

                The firewall was certainly a contributing factor but it didn't solve my root problem.

                 

                This is the error I get when I start up a Linux slave:

                [Server:server-three-slave-HP] 15:28:26,627 ERROR [com.vantage.capital.service.xml.XMLTranServiceBean] (ajp--192.168.2.12-8259-5) DDF0001,CashTransaction:EXCP:org.hibernate.cache.CacheException: org.hibernate.cache.CacheException: org.infinispan.CacheException: org.jgroups.TimeoutException: timeout sending message to master:server-three-Linux/hibernate; org.hibernate.cache.CacheException: org.hibernate.cache.CacheException: org.infinispan.CacheException: org.jgroups.TimeoutException: timeout sending message to master:server-three-Linux/hibernate; org.hibernate.cache.CacheException: org.infinispan.CacheException: org.jgroups.TimeoutException: timeout sending message to master:server-three-Linux/hibernate; org.infinispan.CacheException: org.jgroups.TimeoutException: timeout sending message to master:server-three-Linux/hibernate

                 

                And this is the master equivalent:

                [Server:server-three-Linux] 15:30:27,614 ERROR [com.vantage.capital.service.xml.XMLTranServiceBean] (ajp--192.168.2.3-8259-123) DDF0002,CashWithdrawal:EXCP:org.hibernate.cache.CacheException: org.hibernate.cache.CacheException: org.infinispan.CacheException: org.jgroups.TimeoutException: timeout sending message to slave2:server-three-slave-HP/hibernate; org.hibernate.cache.CacheException: org.hibernate.cache.CacheException: org.infinispan.CacheException: org.jgroups.TimeoutException: timeout sending message to slave2:server-three-slave-HP/hibernate; org.hibernate.cache.CacheException: org.infinispan.CacheException: org.jgroups.TimeoutException: timeout sending message to slave2:server-three-slave-HP/hibernate; org.infinispan.CacheException: org.jgroups.TimeoutException: timeout sending message to slave2:server-three-slave-HP/hibernate

                 

                I get the above errors under very ligh load, I can force the error by holding the enter keydown on two terminals doing transactions (cash deposits to accounts with different currencies, user and branch to eliminate database locking)

                 

                When this happens I see a lot of locking in the Oracle database.

                The DBA (very experiences no ping pong guy) assures me that it is not a database problem.

                In fact I see similar locking activity when I run the master only and that runs very fast.

                 

                I would attach my domain.xml etc but I can't see where I can add an attachment.

                 

                Maybe this is relevant, I use profile "full-ha"

                 

                 

                                    <broadcast-groups>

                                        <broadcast-group name="bg-group1">

                                            <local-bind-port>9877</local-bind-port>

                                            <group-address>231.7.7.7</group-address>

                                            <group-port>9876</group-port>

                                            <broadcast-period>5000</broadcast-period>

                                            <connector-ref>

                                                netty

                                            </connector-ref>

                                        </broadcast-group>

                                    </broadcast-groups>

                 

                                    <discovery-groups>

                                        <discovery-group name="dg-group1">

                                            <group-address>231.7.7.7</group-address>

                                            <group-port>9876</group-port>

                                            <refresh-timeout>10000</refresh-timeout>

                                        </discovery-group>

                                    </discovery-groups>

                 

                                    <cluster-connections>

                                        <cluster-connection name="my-cluster">

                                            <address>jms</address>

                                            <connector-ref>netty</connector-ref>

                                            <discovery-group-ref discovery-group-name="dg-group1"/>

                                        </cluster-connection>

                                    </cluster-connections>

                 

                                    <security-settings>

                                        <security-setting match="#">

                                            <permission type="send" roles="guest"/>

                                            <permission type="consume" roles="guest"/>

                                            <permission type="createNonDurableQueue" roles="guest"/>

                                            <permission type="deleteNonDurableQueue" roles="guest"/>

                                        </security-setting>

                                    </security-settings>

                • 35. Re: mod_cluster 1.2 + JBossAS 7.1 problem
                  ranga033

                  May I request you to start a new thread under AS7 as this is not related to mod_cluster. This would enable more experts from AS7 community to have a look at your issue.

                   

                  Also, the above portion which you have pasted seems to be of messaging subsystem. From your exception it looks like there is a problem with Hibernate cache involving Infinispan/JGroups. This could be because the cache is not yet started or not reachable.

                  1 2 3 Previous Next