Selecting tracing by transaction
For each transaction, you can specify whether standard tracing or special tracing is to be done, or whether tracing is to be suppressed for that transaction altogether.
About this task
For each component, you can specify two sets of trace level attributes. The trace level attributes define the trace point IDs to be traced for that component when standard task tracing is being done and when special task tracing is being done, respectively.
If you are running a test region, you probably have background tracing most of the time. In this case, the default tracing options (standard tracing for all transactions, and level-1 trace points only in the standard set for all components) probably suffice. All you need do is to enable the required trace destinations and set up any related tracing options. Details are given in Setting trace destinations and tracing status.
For a production system, background tracing might incur an unacceptable processing overhead. If you find this to be so, you are recommended to set up tracing so that exception traces only are recorded on an auxiliary trace data set. There need be no other tracing overhead, and you can be sure that the exception trace will be preserved even when the event invoking the trace does not cause a system dump to be taken. For details, see CICS exception tracing.
When specific problems arise, you can set up special tracing so you can focus on just the relevant tasks and components. Use this procedure to specify the tracing you need: