Develop an application to use JMS

This topic gives an overview of the steps needed to develop a J2EE application (servlet or enterprise bean) to use the JMS API directly for asynchronous messaging.

This topic only describes the JMS-related considerations; it does not describe general J2EE application programming, which already be familiar with. For detailed information about these steps, and for examples of developing a J2EE application to use JMS, see Java Message Service Documentation Link outside Information Center and WebSphere MQ Using Java Link outside Information Center.

Details of JMS resources that are used by J2EE applications are defined to WAS and bound into the JNDI namespace by the WebSphere administrative support.

To use JMS, any method of a J2EE application completes the following general steps:

  1. Import JMS packages.

    A J2EE application that uses JMS starts with a number of import statements for JMS, which should include at least the following:

    import javax.jms.*;       //JMS interfaces 
import javax.naming.*;    //Used for JNDI lookup of administered objects

  2. Get an initial context.

       try     {
        ctx = new InitialContext(env);
...

  3. Retrieve administered objects from the JNDI namespace.

    The InitialContext.lookup() method is used to retrieve administered objects (a queue connection factory and the queue destinations); for example, to receive a message from a queue:

           qcf = (QueueConnectionFactory)ctx.lookup( qcfName );
...
       inQueue = (Queue)ctx.lookup( qnameIn );
...

  4. Create a connection to the messaging service provider.

    The connection provides access to the underlying transport, and is used to create sessions. The createQueueConnection() method on the factory object is used to create the connection. 

           connection = qcf.createQueueConnection();

    The JMS specification defines that connections should be created in the stopped state. Until the connection starts, MessageConsumers that are associated with the connection cannot receive any messages. To start the connection, issue the following command:

           connection.start();

  5. Create a session, for sending or receiving messages.

    The session provides a context for producing and consuming messages, including the methods used to create MessageProducers and MessageConsumers. The createQueueSession method is used on the connection to obtain a session. The method takes two parameters:

    • A boolean that determines whether or not the session is transacted.

    • A parameter that determines the acknowledge mode.

           boolean transacted = false;
       session = connection.createQueueSession( transacted,
                                               Session.AUTO_ACKNOWLEDGE);

    In this example, the session is not transacted, and it should automatically acknowledge received messages. With these settings, a message is backed out only after a system error or if the application terminates unexpectedly.

    The following points, as defined in the EJB specification, apply to these flags:

    • The transacted flag passed on createQueueSession is ignored inside a global transaction and all work is performed as part of the transaction. Outside of a transaction the transacted flag is not used and, if set to true, the application should use session.commit() and session.rollback() to control the completion of the work. In an EJB2.0 module, if the transacted flag is set to true and outside of an XA transaction, then the session is involved in the WebSphere local transaction and the unresolved action attribute of the method applies to the JMS work.

    • Clients cannot use using Message.acknowledge() to acknowledge messages. If a value of CLIENT_ACKNOWLEDGE is passed on the createxxxSession call, then messages are automatically acknowledged by the application server and Message.acknowledge() is not used.

  6. Send a message.

    1. Create MessageProducers to create messages.

      For point-to-point messaging the MessageProducer is a QueueSender that is created by passing an output queue object (retrieved earlier) into the createSender method on the session. A QueueSender is normally created for a specific queue, so that all messages sent using that sender are sent to the same destination.

             QueueSender queueSender = session.createSender(inQueue);

    2. Send the message.

      To send the message, the message is passed to the send method on the QueueSender:

             queueSender.send(outMessage);

  7. Receive replies.

    1. Create a correlation ID to link the message sent with any replies.

      In this example, the client receives reply messages that are related to the message that it has sent, by using a provider-specific message ID in a JMSCorrelationID.

             messageID = outMessage.getJMSMessageID();

      The correlation ID is then used in a message selector, to select only messages that have that ID:

             String selector = "JMSCorrelationID = '"+messageID+"'";

    2. Create a MessageReceiver to receive messages.

      For point-to-point the MessageReceiver is a QueueReceiver that is created by passing an input queue object (retrieved earlier) and the message selector into the createReceiver method on the session.

        QueueReceiver queueReceiver = session.createReceiver(outQueue, selector);

    3. Retrieve the reply message.

      To retrieve a reply message, the receive method on the QueueReceiver is used:

        Message inMessage = queueReceiver.receive(2000);

      The parameter in the receive call is a timeout in milliseconds. This parameter defines how long the method should wait if there is no message available immediately. If you omit this parameter, the call blocks indefinitely. If you do not want any delay, use the receiveNoWait()method. In this example, the receive call returns when the message arrives, or after 2000ms, whichever is sooner.

    4. Act on the message received.

      When a message is received, you can act on it as needed by the business logic of the client. Some general JMS actions are to check that the message is of the correct type and extract the content of the message. To extract the content from the body of the message, it is necessary to cast from the generic Message class (which is the declared return type of the receive methods) to the more specific subclass, such as TextMessage. It is good practice always to test the message class before casting, so that unexpected errors can be handled gracefully.

      In this example, the instanceof operator is used to check that the message received is of the TextMessage type. The message content is then extracted by casting ti the TextMessage subclass.

             if ( inMessage instanceof TextMessage )

...
           String replyString = ((TextMessage) inMessage).getText();

  8. Closing down.

    If the application needs to create many short-lived JMS objects at the Session level or lower, it is important to close all the JMS resources used. To do this, you call the close() method on the various classes (QueueConnection, QueueSession, QueueSender, and QueueReceiver) when the resources are no longer required.

           queueReceiver.close();
...
       queueSender.close();
...
       session.close();
       session = null;
...
       connection.close();
       connection = null;

  9. Publishing and subscribing messages.

    To use JMS Publish/Subscribe support instead of point-to-point messaging, the general actions are the same; for example, to create a session and connection. The exceptions are that topic resources are used instead of queue resources (such as TopicPublisher instead of QueueSender), as shown in the following example to publish a message:

    // Creating a TopicPublisher
       TopicPublisher pub = session.createPublisher(topic);
...
       pub.publish(outMessage);
...
       // Closing TopicPublisher
       pub.close();

  10. Handling errors

    Any JMS runtime errors are reported by exceptions. The majority of methods in JMS throw JMSExceptions to indicate errors. It is good programming practice to catch these exceptions and display them on a suitable output.

    Unlike normal Java exceptions, a JMSException can contain another exception embedded in it. The implementation of JMSException does not include the embedded exception in the output of its toString()method. Therefore, you need to check explicitly for an embedded exception and print it out, as shown in the following example:

        catch (JMSException je)
    {
       System.out.println("JMS failed with "+je);
       Exception le = je.getLinkedException();
       if (le != null)
       {
           System.out.println("linked exception "+le);
       }
    }