problems with multiple HornetQ instances in the same machine

[f.formizzi]

Hi, I've a problem to create a cluster of multiple instances in the same machine.

Here the steps I followed:

1) I installed HornetQ 2.0.0GA under JBoss 4.2.3GA, launching $HORNETQ_HOME\config\jboss-as\ant as4

2) under $JBOSS_HOME\server I copied the directory "all-with-hornetq"  I renamed in "all-with-hornetq2". (I've 2 server, the first is "all-with-hornetq", the second is "all-with-hornetq2").

3) in the file   $JBOSS_HOME\server\all-with-hornetq\conf\jboss-service.xml I added this:

  <mbean code="org.jboss.services.binding.ServiceBindingManager"
     name="jboss.system:service=ServiceBindingManager">
     <attribute name="ServerName">ports-01</attribute>
     <attribute name="StoreURL">${jboss.home.url}/docs/examples/binding-manager/sample-bindings.xml</attribute>
     <attribute name="StoreFactoryClassName">
       org.jboss.services.binding.XMLServicesStoreFactory
     </attribute>
   </mbean>

4)  I do the same thing in  $JBOSS_HOME\server\all-with-hornetq2\conf\jboss-service.xml, setting ServerName to ports-02:

       <attribute name="ServerName">ports-02</attribute>

5) port-01 and ports-02 are correctly configured in sample-binding.xml (default configuration), so I didn't modify this file.

6) I changed the values of netty-connector and netty-acceptor ports:

     first instance: port 5445

     second instance: port 5545

7) I launched the first instance wiht the command:

     ./run.sh -c all-with-hornetq -b=10.212.150.37 -Djboss.partition.name=mypart -Dhornetq.remoting.netty.host=10.212.15.37

My IP address is obviously 10.212.150.37

8) I launched the second instance wiht the command:

     ./run.sh -c all-with-hornetq2 -b=10.212.150.37 -Djboss.partition.name=vmypart -Dhornetq.remoting.netty.host=10.212.15.37

The second instance doesn't work, and I've a lot of this log message:

WARN  [DiscoveryGroupImpl] There seem to be more than one broadcasters on the network broadcasting the same node id

If I launch only one instance everything works perfectly. No problems if the instance is the second.

If I do the same operations in two different machines, everything works perfectly (without step 2-3-4-5-6, I changed only ip address).

Where I wrong??

[timfox]

Since both "nodes" are on the same machine, each one is probably pointing at the same persistent store - this won't work, they will be writing over and reading each others data.

If you want to do this, you need to configure each node to have different storage.

Take a look at the clustering examples in the distro - they configure different storage for different nodes.

[f.formizzi]

Thanks for your reply, It was useful to solve the problem!

If anyone has the same problem, as Tim said, you must add the follow directive in starting script:

-Ddata.dir=../data-server1 (for example in node 1)

-Ddata.dir=../data-server2 (for example in node 2)

easy, not?

[dembuongactro]

Thanks,

Francesco