Message tracing

There are many situations where it may be useful for you to trace a message’s path through the engine. For example, when troubleshooting a message that is not going where expected, or debugging a route that is not behaving as expected.

Message tracing can:

  • Let you see what messages enter the engine and the resultant messages that go out.
  • Take an outbound message and see from what inbound message it derived.
  • Take a message and see what reply messages it generated.
  • Take a reply message and see what message to which it is in reply.

To trace messages that go through the engine, you must record information each time a message ID changing action happens. Information is recorded every time a message with a new message ID is created, a message ID is reassigned, or a message is stopped or outputted.

If only the message ID and the source message ID are recorded at each of these points, then you cannot trace where each message came from and is going.

After this information is recorded, a command-line interface takes the message ID of an inbound message and returns the message IDs of all outbound messages resulting from it. It takes the message ID of an outbound message and returns the message ID of the message where it ultimately originated.

Inter-site routing requires that a special additional tracing record be recorded on the destination site.

See Inter-site route.