Sometimes a problem appears in a message flow that has previously run successfully. To identify the cause of the problem, answer the following questions:
If so, it is likely that the error exists somewhere in the new or modified part of the flow. Examine the changes and see if you can find an obvious reason for the problem.
Did the problem occur when you used part of the message flow that had never been invoked before? If so, it is likely that the error exists in that part. Try to find out what the message flow was doing when it failed by using user tracing.
If you have run a message flow successfully on many previous occasions, check the current queue status and the files that were being processed when the error occurred. It is possible that they contain some unusual data value that invokes a rarely-used path in the message flow.
If a message with an unexpected message format has been put onto a queue (for example, a message from a queue manager on a different operating system) it might require data conversion or a different form of processing. Also, check whether you have changed any of the message formats that are used.
Is there something different about the way that your system is set up that is causing the problem? For example, have the queues been defined with the same maximum message length or priority? Are there differences in the databases used, or their setup?
Your message flow might be losing errors because of incorrect use of the failure terminals on built-in nodes. If you use the failure terminals, make sure that you handle errors adequately. See Handling errors in message flows for more information about failure terminals.