MQSeries.net :: View topic

BarryB · Posted: Wed Jul 28, 2004 5:22 am Post subject:

I have the following scenario:

Box name = MQCluster1 (resides on AIX)
On this box there is a ServConn Channel called MQCluster1.SRV.TCP.ALL
Also, there is a queue called ROCC.RECV.QUEUE which is a cluster queue local on MQCluster1

This is a production box.

There is an MQClient application on the following box:

Box name = DALLAS (resides on AIX)
This application performs a 'get with wait unlimited' on the ROCC.RECV.QUEUE cluster queue, which is local on MQCluster1.

Now, there is a 'training' application on the following box:

Box name = TRAINING (resides on AIX)
The training application 'puts' a message to a remote queue definition on MQCluster1, which is transported to the real local queue on the mainframe.

Both the training and the MQClient application mentioned above specify the same channel name to connect to MQCluster1. Everything works perfectly with the MQClient application performing a get with wait unlimited until the training application puts a message on the remote queue using the same channel name as the MQClient app. Once the training app puts a message, if there is any incoming messages for the MQClient app to get, the MQClient app just hangs and never gets a message.

Is there a possibility that because the training and MQClient apps specify the same channel, that there could be some contention???

Any ideas would be appreciated. Also, the training app is not a part of the MQCluster, and is using a remote queue which is obviously not a cluster queue.

Thanks,

B