HornetQ 2.3.0 Alpha replication

[amshaik]

Hi,

I'm attempting to configure replication and failover in HornetQ 2.3.0 Alpha using static connectors. I'm able to start the live server, but when I attempt to start the backup server, it hangs for some time, and then I receive the following error:

ERROR: HQ114002: Failure in initialisation

java.lang.RuntimeException: Could not estabilish the connection

        at org.hornetq.core.server.impl.HornetQServerImpl$SharedNothingBackupActivation.run(HornetQServerImpl.java:2152)

        at java.lang.Thread.run(Thread.java:662)

java.lang.RuntimeException: Could not estabilish the connection

        at org.hornetq.core.server.impl.HornetQServerImpl$SharedNothingBackupActivation.run(HornetQServerImpl.java:2152)

        at java.lang.Thread.run(Thread.java:662)

Aug 8, 2012 10:08:12 AM org.hornetq.integration.bootstrap.HornetQBootstrapServer shutDown

INFO: HQ101002: Stopping HornetQ Server

In my live server's hornetq-configuration.xml, I have:

<configuration xmlns="urn:hornetq"

               xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance"

               xsi:schemaLocation="urn:hornetq /schema/hornetq-configuration.xsd">

   <bindings-directory>${data.dir:../../data}/bindings</bindings-directory>

   <journal-directory>${data.dir:../../data}/journal</journal-directory>

   <large-messages-directory>${data.dir:../../data}/large-messages</large-messages-directory>

  <connectors>

      <connector name="netty-backup">

        <factory-class>org.hornetq.core.remoting.impl.netty.NettyConnectorFactory</factory-class>

        <param key="host"  value="${hornetq.remoting.netty.host: myBackupHost}"/>

        <param key="port"  value="${hornetq.remoting.netty.port:5445}"/>

      </connector>

  </connectors>

  <acceptors>

      <acceptor name="netty">

         <factory-class>org.hornetq.core.remoting.impl.netty.NettyAcceptorFactory</factory-class>

         <param key="host"  value="${hornetq.remoting.netty.host:0.0.0.0}"/>

         <param key="port"  value="${hornetq.remoting.netty.port:5445}"/>

      </acceptor>

  </acceptors>

  <security-settings>

      <security-setting match="#">

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

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

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

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

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

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

      </security-setting>

  </security-settings>

  <shared-store>false</shared-store>

  <clustered>true</clustered>

  <allow-failback>true</allow-failback>

  <cluster-connections>

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

                <address>jms</address>

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

                <static-connectors allow-direct-connections-only="true">

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

                </static-connectors>

        </cluster-connection>

  </cluster-connections>

</configuration>

My backup server's configuration is:

<configuration xmlns="urn:hornetq"

               xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance"

               xsi:schemaLocation="urn:hornetq /schema/hornetq-configuration.xsd">

  <backup>true</backup>

  <shared-store>false</shared-store>

  <live-connector-ref connector-name = "netty-live"/>

  <clustered>true</clustered>

  <connectors>

      <connector name = "netty-live">

        <factory-class>org.hornetq.core.remoting.impl.netty.NettyConnectorFactory</factory-class>

        <param key="host"  value="${hornetq.remoting.netty.host:myLiveHost}"/>

        <param key="port"  value="${hornetq.remoting.netty.port:5445}"/>

      </connector>

   </connectors>

   <acceptors>

      <acceptor name="netty">

         <factory-class>org.hornetq.core.remoting.impl.netty.NettyAcceptorFactory</factory-class>

         <param key="host"  value="${hornetq.remoting.netty.host:0.0.0.0}"/>

         <param key="port"  value="${hornetq.remoting.netty.port:5445}"/>

      </acceptor>

   </acceptors>

   <cluster-connections>

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

                <address>jms</address>

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

                <retry-interval>500</retry-interval>

                <reconnect-attempts>-1</reconnect-attempts>

                <static-connectors allow-direct-connections-only="true">

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

                </static-connectors>

        </cluster-connection>

   </cluster-connections>

   <paging-directory>${data.dir:../../data1}/paging</paging-directory>

   <bindings-directory>${data.dir:../../data1}/bindings</bindings-directory>

   <journal-directory>${data.dir:../../data1}/journal</journal-directory>

   <large-messages-directory>${data.dir:../../data1}/large-messages</large-messages-directory>

   <security-settings>

      <security-setting match="#">

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

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

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

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

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

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

      </security-setting>

   </security-settings>

</configuration>

Can anybody pinpoint what's going wrong here?

[borges]

I believe you are missing a connector to itself at the backup.

There is an example at "examples/jms/replicated-transaction-failover".

Let me know if that works or not.

FWIW, the spelling error at the exception message was fixed some weeks ago.

[amshaik]

Thanks Francisco. I added a connector to itself in both the live and back configs, and it worked.

[amshaik]

But I'm still wondering, why exactly do the servers need a connector to themselves? How is it actually used?

[clebert.suconic]

The backup needs to know how to connect to the live server.

There's some configuration being changed now by Andy Taylor.

Any feedback on usuability is appreciated.

[amshaik]

Thanks. My next step is to attempt to configure failback. I set <allow-failback> to true. And then, I noticed that the example uses a "failback-connector" element. For example:

<failback-connector>

         <factory-class>org.hornetq.core.remoting.impl.netty.NettyConnectorFactory</factory-class>

         <param key="host"  value="${hornetq.remoting.netty.host:myBackupHost}"/>

         <param key="port"  value="${hornetq.remoting.netty.port:5446}"/>

</failback-connector>

This is where I ran into trouble. Do I still need to specify a  <static-connector> in the <cluster-connection> tag?

If so, how do I reference the failback-connector, given that it doesn't have a name attribute?

[borges]

0. While we very much appreciate you testing the waters of the Alpha, I need to warn you that the fail-back configuration and the backup configuration used in this alpha release were found to be not flexible enough, and were changed in our master repository (the merge took place today).

On your specific questions, IIRC,

1. You still need the previous configuration as it was, but now you also need the fail-back connector.

2. Unlike the other connector's this one does not have a name attribute.

[wporter]

Does the fail-back connector reside in the live server configuration only?  Does this need to be referenced in the cluster configuration or does just declaring it in the configuration section suffice?