ACK after closing the consumer

[clebert.suconic]

I know this is illegal:

         consumer.close();

         for (ClientMessage ack : myMessages)

         {

            ack.acknowledge();

         }

  //consumerNonPaged.close();

But shouldn't we throw an exception on the client preventing it?

Right now this will cause a NPE at the server.

Low priority issue.. I will create the JIRA. I'm just raising it here.

[Thread-0 (group:HornetQ-remoting-threads88416491-2104764160)] 17:20:41,638 SEVERE [org.hornetq.core.protocol.core.ServerSessionPacketHandler]  Caught unexpected exception

java.lang.NullPointerException

at org.hornetq.core.server.impl.ServerSessionImpl.individualAcknowledge(ServerSessionImpl.java:525)

at org.hornetq.core.protocol.core.ServerSessionPacketHandler.handlePacket(ServerSessionPacketHandler.java:436)

at org.hornetq.core.protocol.core.impl.ChannelImpl.handlePacket(ChannelImpl.java:471)

at org.hornetq.core.protocol.core.impl.RemotingConnectionImpl.doBufferReceived(RemotingConnectionImpl.java:451)

at org.hornetq.core.protocol.core.impl.RemotingConnectionImpl.bufferReceived(RemotingConnectionImpl.java:412)

at org.hornetq.core.remoting.server.impl.RemotingServiceImpl$DelegatingBufferHandler.bufferReceived(RemotingServiceImpl.java:459)

at org.hornetq.core.remoting.impl.invm.InVMConnection$1.run(InVMConnection.java:135)

at org.hornetq.utils.OrderedExecutorFactory$OrderedExecutor$1.run(OrderedExecutorFactory.java:100)

at java.util.concurrent.ThreadPoolExecutor$Worker.runTask(ThreadPoolExecutor.java:886)

at java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:908)

at java.lang.Thread.run(Thread.java:619)

[timfox]

Are you sure this is illegal? With JMS CLIENT_ACKNOWLEDGE it is legal to acknowledge messages after the consumer is closed.

[clebert.suconic]

if it's legal, there's a bug then

[clebert.suconic]

Tim: Are you sure it would be legal to do that?

Currently the delivering refs is only controlled at the ServerConsumer.  All the references are sent back to the queue when the consumer is closed.

If we want to allow this, the whole delivering references should have to be refactored to the session and the ACK to be controlled at the Session (and not at the Consumer)

[clebert.suconic]

Actually, I was looking at JBoss Messaging's code. The delivering references are controlled at the Session and not at the consumer.

I will do the refactoring, unless you have anything to say.

[timfox]

Personally, I would leave it for now, unless a user complains, it's not a huge deal.