BeanEndpoint can receive messages before it has fully started, causing 
exceptions 
----------------------------------------------------------------------------------

                 Key: SM-1171
                 URL: https://issues.apache.org/activemq/browse/SM-1171
             Project: ServiceMix
          Issue Type: Bug
    Affects Versions: 3.2.1, 3.2
            Reporter: Ryan Bohn


I have a Sender endpoint and a Receiver endpoint, both BeanEndpoint objects.  
The Sender endpoint immediately starts sending messages.  Every so often, the 
receiver endpoint accepts tries to process a message the start() method has 
completed, causing the following exception:

java.lang.NullPointerException
        at 
org.apache.servicemix.bean.BeanEndpoint.getCorrelationExpression(BeanEndpoint.java:421)
        at 
org.apache.servicemix.bean.BeanEndpoint.getCorrelation(BeanEndpoint.java:298)
        at 
org.apache.servicemix.bean.BeanEndpoint.onProviderExchange(BeanEndpoint.java:218)
        at 
org.apache.servicemix.bean.BeanEndpoint.process(BeanEndpoint.java:211)
        at 
org.apache.servicemix.common.AsyncBaseLifeCycle.doProcess(AsyncBaseLifeCycle.java:538)
        at 
org.apache.servicemix.common.AsyncBaseLifeCycle.processExchange(AsyncBaseLifeCycle.java:490)
        at 
org.apache.servicemix.common.BaseLifeCycle.onMessageExchange(BaseLifeCycle.java:46)
        at 
org.apache.servicemix.jbi.messaging.DeliveryChannelImpl.processInBound(DeliveryChannelImpl.java:610)
        at 
org.apache.servicemix.jbi.nmr.flow.AbstractFlow.doRouting(AbstractFlow.java:170)
        at 
org.apache.servicemix.jbi.nmr.flow.seda.SedaFlow.doRouting(SedaFlow.java:167)
        at 
org.apache.servicemix.jbi.nmr.flow.seda.SedaQueue$1.run(SedaQueue.java:134)
        at 
java.util.concurrent.ThreadPoolExecutor$Worker.runTask(ThreadPoolExecutor.java:650)
        at 
java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:675)
        at java.lang.Thread.run(Thread.java:613)

For now, I've put a sleep in the sender endpoint, which fixes the problem.  
Obviously, this a hack.

I originally thought that the original solution would be to move the 
super.start() call to the end of the start() method.  However, this turned out 
to cause problems for the sender endpoint.

I have two options to propose:

- use synchronization to ensure the bean has started, so that the process() 
method blocks until the start() method has finished.  This has the disadvantage 
of incurring a synchronization hit on each message exchange.

- separate the activation of an endpoint from the registration.  This way an 
endpoint would be ready to go by being activated, but wouldn't be visible to 
other endpoints until the start() method has finished.  I'm not sure what code 
changes would need be necessary to make this happen.



-- 
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.

Reply via email to