5 Replies Latest reply on Sep 18, 2002 9:26 AM by Ciprian

    farming

    seanxu Newbie

      I copy my ear and war files to server2's deploy/farm directory but it does not deployed to the server, server1.

      I have two JBoss servers intalled on different hosts on the same local network.
      I start server1 first and when its startup completely done, I start server2. I got the following message in the log:
      12:37:27,504 INFO [DefaultPartition:ReplicantManager] in membershipChanged
      12:37:27,505 INFO [DefaultPartition:ReplicantManager] deadMembers: 0
      12:37:27,505 INFO [DefaultPartition:ReplicantManager] There are new members. S
      pawning MergeMembers thread.
      12:37:27,973 WARN [DefaultPartition:ReplicantManager] mergeMembers received NUL
      L from lookupLocalReplicants
      But it does not report server1 got the deployment from server2. Further test using browser to access the test url get the error saying the context is not created under server1.

      Is there anything I did wrong?

      I know the ejb does not have clustered enabled but other war files for servlet and htmls should be deployed.

      thanks

        • 1. Re: farming
          Dennis Cartier Newbie

          I am having the same trouble. I even created the farm-service.xml file that the Quick Start Guide walks you through creating.

          No deployment takes place on the other members of the cluster.

          Anyone have this working?

          I am using 3.0.1RC1 running the all config by the way.

          Dennis

          • 2. Re: farming
            Tung Dang Novice

            I have the same problem, as well.

            Can any one give me any clues?

            Regards.

            • 3. Re: farming
              Ciprian Newbie

              hi,

              I have a cluster with 2 machines I succesfuly deployed a stateful/entity bean in farming. I deploy it on one machine in /farm and was replicated in entire cluster.

              1.
              Make sure you specified
              True
              in jboss.xml of your bean.

              2.
              Also I have the following farm-service.xml in server/all/deploy folder:

              <?xml version="1.0" encoding="UTF-8"?>



              jboss:service=DefaultPartition
              DefaultPartition
              ./farm
              jboss.deployment:type=DeploymentScanner,flavor=URL



              3.
              My cluster is Manualy configured over TCP in cluster-service.xml. I couldn't set it up with automatic node discovery.

              On machine 1:
              TCP(start_port=7800):TCPPING(initial_hosts=IP_OF_YOUR_MACHINE_2[7800];port_range=5;timeout=3000;num_initial_members=2;up_thread=true;down_thread=true):VERIFY_SUSPECT(timeout=1500;down_thread=false;up_thread=false):pbcast.STABLE(desired_avg_gossip=20000;down_thread=false;up_thread=false):pbcast.NAKACK(down_thread=true;up_thread=true;gc_lag=100;retransmit_timeout=3000):pbcast.GMS(join_timeout=5000;join_retry_timeout=2000;shun=false;print_local_addr=false;down_thread=true;up_thread=true)
              On machine 2:
              TCP(start_port=7800):TCPPING(initial_hosts=IP_OF_YOUR_MACHINE_1[7800];port_range=5;timeout=3000;num_initial_members=2;up_thread=true;down_thread=true):VERIFY_SUSPECT(timeout=1500;down_thread=false;up_thread=false):pbcast.STABLE(desired_avg_gossip=20000;down_thread=false;up_thread=false):pbcast.NAKACK(down_thread=true;up_thread=true;gc_lag=100;retransmit_timeout=3000):pbcast.GMS(join_timeout=5000;join_retry_timeout=2000;shun=false;print_local_addr=false;down_thread=true;up_thread=true)

              4.
              Sometimes deployment in farming it doesn't work to me too.
              That's happening after some wrong deployment configurations (when getting some exceptions in servers).
              The only solutions was to restart the entire cluster (all machines).

              5.
              Anyway I get a strange behavior:
              I have an entity bean deployed in farming on both nodes in the cluster (with clustering activated)
              I am accesing it from a remote client and add 3 records in the table. The records are inserted: one on machine1 and two records on machine 2 (load-balancing). Now when I am using the default implementation of findAll() to get all records I get only the ones from one machine not all three records.

              Can somebody gime me a hint why that's happening?

              • 4. Re: farming
                Ciprian Newbie

                hi,

                I have a cluster with 2 machines I succesfuly deployed a stateful/entity bean in farming. I deploy it on one machine in /farm and was replicated in entire cluster.

                1.
                Make sure you specified
                True
                in jboss.xml of your bean.

                2.
                Also I have the following farm-service.xml in server/all/deploy folder:

                <?xml version="1.0" encoding="UTF-8"?>



                jboss:service=DefaultPartition
                DefaultPartition
                ./farm
                jboss.deployment:type=DeploymentScanner,flavor=URL



                3.
                My cluster is Manualy configured over TCP in cluster-service.xml. I couldn't set it up with automatic node discovery.

                On machine 1:
                TCP(start_port=7800):TCPPING(initial_hosts=IP_OF_YOUR_MACHINE_2[7800];port_range=5;timeout=3000;num_initial_members=2;up_thread=true;down_thread=true):VERIFY_SUSPECT(timeout=1500;down_thread=false;up_thread=false):pbcast.STABLE(desired_avg_gossip=20000;down_thread=false;up_thread=false):pbcast.NAKACK(down_thread=true;up_thread=true;gc_lag=100;retransmit_timeout=3000):pbcast.GMS(join_timeout=5000;join_retry_timeout=2000;shun=false;print_local_addr=false;down_thread=true;up_thread=true)
                On machine 2:
                TCP(start_port=7800):TCPPING(initial_hosts=IP_OF_YOUR_MACHINE_1[7800];port_range=5;timeout=3000;num_initial_members=2;up_thread=true;down_thread=true):VERIFY_SUSPECT(timeout=1500;down_thread=false;up_thread=false):pbcast.STABLE(desired_avg_gossip=20000;down_thread=false;up_thread=false):pbcast.NAKACK(down_thread=true;up_thread=true;gc_lag=100;retransmit_timeout=3000):pbcast.GMS(join_timeout=5000;join_retry_timeout=2000;shun=false;print_local_addr=false;down_thread=true;up_thread=true)

                4.
                Sometimes deployment in farming it doesn't work to me too.
                That's happening after some wrong deployment configurations (when getting some exceptions in servers).
                The only solutions was to restart the entire cluster (all machines).

                5.
                Anyway I get a strange behavior:
                I have an entity bean deployed in farming on both nodes in the cluster (with clustering activated)
                I am accesing it from a remote client and add 3 records in the table. The records are inserted: one on machine1 and two records on machine 2 (load-balancing). Now when I am using the default implementation of findAll() to get all records I get only the ones from one machine not all three records.

                Can somebody gime me a hint why that's happening?

                • 5. Re: farming
                  Ciprian Newbie

                  hi,

                  I have a cluster with 2 machines I succesfuly deployed a stateful/entity bean in farming. I deploy it on one machine in /farm and was replicated in entire cluster.

                  1.
                  Make sure you specified
                  True
                  in jboss.xml of your bean.

                  2.
                  Also I have the following farm-service.xml in server/all/deploy folder:

                  <?xml version="1.0" encoding="UTF-8"?>



                  jboss:service=DefaultPartition
                  DefaultPartition
                  ./farm
                  jboss.deployment:type=DeploymentScanner,flavor=URL



                  3.
                  My cluster is Manualy configured over TCP in cluster-service.xml. I couldn't set it up with automatic node discovery.

                  On machine 1:
                  TCP(start_port=7800):TCPPING(initial_hosts=IP_OF_YOUR_MACHINE_2[7800];port_range=5;timeout=3000;num_initial_members=2;up_thread=true;down_thread=true):VERIFY_SUSPECT(timeout=1500;down_thread=false;up_thread=false):pbcast.STABLE(desired_avg_gossip=20000;down_thread=false;up_thread=false):pbcast.NAKACK(down_thread=true;up_thread=true;gc_lag=100;retransmit_timeout=3000):pbcast.GMS(join_timeout=5000;join_retry_timeout=2000;shun=false;print_local_addr=false;down_thread=true;up_thread=true)
                  On machine 2:
                  TCP(start_port=7800):TCPPING(initial_hosts=IP_OF_YOUR_MACHINE_1[7800];port_range=5;timeout=3000;num_initial_members=2;up_thread=true;down_thread=true):VERIFY_SUSPECT(timeout=1500;down_thread=false;up_thread=false):pbcast.STABLE(desired_avg_gossip=20000;down_thread=false;up_thread=false):pbcast.NAKACK(down_thread=true;up_thread=true;gc_lag=100;retransmit_timeout=3000):pbcast.GMS(join_timeout=5000;join_retry_timeout=2000;shun=false;print_local_addr=false;down_thread=true;up_thread=true)

                  4.
                  Sometimes deployment in farming it doesn't work to me too.
                  That's happening after some wrong deployment configurations (when getting some exceptions in servers).
                  The only solutions was to restart the entire cluster (all machines).

                  5.
                  Anyway I get a strange behavior:
                  I have an entity bean deployed in farming on both nodes in the cluster (with clustering activated)
                  I am accesing it from a remote client and add 3 records in the table. The records are inserted: one on machine1 and two records on machine 2 (load-balancing). Now when I am using the default implementation of findAll() to get all records I get only the ones from one machine not all three records.

                  Can somebody gime me a hint why that's happening?