Debugging by adding Trace nodes to a message flow

You must add a Trace node when the message flow is designed. Viewing the logical message tree in trace output explains how to view the structure of the logical message tree at any point in the message flow, and contains an example of the message content. You can turn the Trace node off when a message flow is promoted to production to improve performance, but you can turn the node on when required. Performance can be affected when Trace nodes are active. The extent to which performance is affected depends on the destination that you choose for the debugging messages; for example, writing to user trace is typically faster than writing to a file or to the system log.

The debug messages can include writing part or all of the logical message tree, but they can also include hard-coded strings to identify a particular point in the message flow (such as PRINTF in a C program). If you write the entire message tree in a Trace node, the behavior of the message flow might be changed. Typically, only the parts of the message that are referenced are parsed, rather than the entire message.