Creating JMS client producer sending to cluster.

[raymanf]

Hi,

I am using HornetQ 2.1.1 integrated inside Spring3.1 (stand alone env)

I have managed finally to configure my cluster mode(HA). I can confirm this when the backup server comes up it writes:

Feb 25, 2013 9:20:07 AM org.hornetq.core.logging.impl.JULLogDelegate info
INFO: backup announced

I have created a client in order to test my env. my target is:

Sending a message after taking down the master cluster node. And have the backup node to deal with the request without the client being noticed.

When I used to work in jboss I could set a Partition Name/Discoverygroup so we dont need to set an exact target ip of a node in the client producder (for such cases).

But now that I work in standalone env I have no idea how to do it.

How should I modified my client and my cluster env code in order to send to Partition name/DiscoveryGroup and not an exact ip a jms message?

so when 1 node being off the second one will take the request automaticlly (after all that is what my cluster about:) )

My client produder code:

Properties p = new Properties();
                              p.put(Context.INITIAL_CONTEXT_FACTORY, "org.jnp.interfaces.NamingContextFactory");
                              p.put(Context.URL_PKG_PREFIXES, "jboss.naming:org.jnp.interfaces");
                              p.put(Context.PROVIDER_URL, "localhost:1099");
                              InitialContext context = new InitialContext(p);
                              TopicConnectionFactory myConnFactory;
                              Topic myTopic;
                              System.out.println("connection factory lookup");
                              myConnFactory = (javax.jms.TopicConnectionFactory) context.lookup("java:/ConnectionFactory");
                              System.out.println("topic lookup");
                              myTopic = (Topic) context.lookup("/topic/liquiditiyFeedTopic");
                              topicConnection = myConnFactory.createTopicConnection();
                              topicSession = topicConnection.createSession(false, Session.AUTO_ACKNOWLEDGE);
                                topicConnection.start();

                                                       

thanks,

ray.

[raymanf]

I understood that I need to use DiscoveryGroupConfiguration.

But cant find some example how to use it.

any direction?

thanks.

[gaohoward]

I would suggest you take a look at some of HornetQ's cluster/failover examples, such as non-transactional-failover example.

[raymanf]

Could you refer me to a link?

thanks.

[raymanf]

I am realy looking for a simple client example. but hard to find.

I am stuck within the DiscoveryGroupConfiguration part:

Connection connection = null;
                    try
                    {

                              Topic topic = HornetQJMSClient.createTopic("liquiditiyFeedTopic");

                              Map<String, Object> params = new HashMap<String, Object>();
                              params.put(TransportConstants.PORT_PROP_NAME, 5445);
                              params.put(TransportConstants.HOST_PROP_NAME, "localhost");
                              TransportConfiguration config = new TransportConfiguration(
                                                  "org.hornetq.core.remoting.impl.netty.NettyConnectorFactory", params);


// no idea how to set it here?---

                              DiscoveryGroupConfiguration dc=new DiscoveryGroupConfiguration(name, localBindAddress, groupAddress, groupPort, refreshTimeout, discoveryInitialWaitTimeout);


// no idea how to set it here?---


                              HornetQConnectionFactory jmsConnectionFactory = HornetQJMSClient.createConnectionFactoryWithHA(groupConfiguration, JMSFactoryType.CF);


                  connection = jmsConnectionFactory.createConnection();


                  session = connection.createSession(false, Session.AUTO_ACKNOWLEDGE);


                  producer = session.createProducer(topic);


                              // Step 7. Create a Text Message


                              // Step 8. Send the Message


TextMessage message = session.createTextMessage(msg);
                    producer.send(message);

mybe you could assit?

thanks.

[gaohoward]

You just need to go to download page http://www.jboss.org/hornetq/downloads and download the latest hornetq distribution, unzip it. All examples are included within the download.

Howard

[ataylor]

have a look at the examples like Howard said, also take a look at the javadocs for HornetQJMSClient

[gaohoward]

You probably need this API?  

public static HornetQConnectionFactory createConnectionFactoryWithHA(final DiscoveryGroupConfiguration groupConfiguration, JMSFactoryType jmsFactoryType);

It's in HornetQJMSClient class.

Howard

[raymanf]

Yes.. already figured that out.. searching now in the examples to look for something similar..

[raymanf]

i was actully searching the whole examples couldnt find any mention to DiscoveryGroupConfiguration. mybe my whole direction is wrong? hope my question was clear.

[raymanf]

Hi,

Ok I have figure it out myself and it seems to work while the two nodes up.

as soon as I kill the first node (to test the cluster) and then try to produce another message from the client I get that exception:

javax.jms.IllegalStateException: Session is closed

          at org.hornetq.jms.client.HornetQSession.checkClosed(HornetQSession.java:1008)

          at org.hornetq.jms.client.HornetQSession.createTextMessage(HornetQSession.java:194)

          at SendMsgToTopicClient.sendMessageToTopic(SendMsgToTopicClient.java:143)

          at SendMsgToTopicClient.main(SendMsgToTopicClient.java:158)

thats my client code:

Topic topic = HornetQJMSClient.createTopic("liquiditiyFeedTopic");
                              Map<String, Object> params = new HashMap<String, Object>();
                              params.put(TransportConstants.PORT_PROP_NAME, 5445);
                              params.put(TransportConstants.HOST_PROP_NAME, "localhost");
DiscoveryGroupConfiguration dc = new DiscoveryGroupConfiguration("231.7.7.7", 9876);
                              HornetQConnectionFactory jmsConnectionFactory = HornetQJMSClient.createConnectionFactoryWithHA(dc, JMSFactoryType.CF);
                              connection = jmsConnectionFactory.createConnection();
                              session = connection.createSession(false, Session.AUTO_ACKNOWLEDGE);
producer = session.createProducer(topic);

* I know the cluster works fine since the second node writes : INFO: backup announced"

thanks.

[gaohoward]

Can you post the complete client code? Please note you should cache your producer for message sending. don't create a new connection/session/producer each time you send a message.

Howard

[raymanf]

I have attached the client code and in other hand the hornetq cluster server configurations of server1 and server2

[raymanf]

Those are the logs from server2:

Feb 26, 2013 9:33:47 AM org.hornetq.core.logging.impl.JULLogDelegate info

INFO: Waiting to become backup node

Feb 26, 2013 9:33:47 AM org.hornetq.core.logging.impl.JULLogDelegate info

INFO: ** got backup lock

Feb 26, 2013 9:33:47 AM org.hornetq.core.logging.impl.JULLogDelegate info

INFO: Using NIO Journal

Feb 26, 2013 9:33:47 AM org.hornetq.core.logging.impl.JULLogDelegate warn

WARNING: Security risk! It has been detected that the cluster admin user and password have not been changed from the installation default. Please see the HornetQ user guide, cluster chapter, for instructions on how to do this.

Feb 26, 2013 9:33:47 AM org.hornetq.core.logging.impl.JULLogDelegate info

INFO: HornetQ Backup Server version 2.2.14.Final (HQ_2_2_14_FINAL, 122) [67c8a76f-7e9b-11e2-9bdf-f5ead70f3c1e] started, waiting live to fail before it gets active

Feb 26, 2013 9:33:48 AM org.hornetq.core.logging.impl.JULLogDelegate info

INFO: backup announced

[raymanf]

I was trying to look for this things in the docs. when the live server falling down does the backup server should write that it got active? coz it doesnt. the only thing I see on the backup startup is:

INFO: backup announced

Is there any other indication that the clusters nodes contacting each other completly?

thanks.