Using trace

You can use trace if you want to find out just how long an individual task waits to be attached to the dispatcher.

About this task

If you do not want to do any other tracing, internal trace is probably a suitable destination for trace entries. Because the task you are interested in is almost inactive, very few trace entries are generated.

Procedure

  1. Select special tracing for the transaction associated with the task, and turn off all standard tracing by setting the master system trace flag off.
  2. Define as special trace points the level-1 trace points for transaction manager (XM), and for the CICS task controlling the facility that initiates the task, such as terminal control (TC). Make sure that no other trace points are defined as special. For guidance about setting up these tracing options, see CICS trace.
  3. When you have selected the options, start tracing to the internal trace table and attempt to initiate the task.
  4. When the task starts, get a system dump using the command CEMT PERFORM SNAP. Format the dump using the keyword TR, to get the internal trace table.
  5. Look for the trace entry showing terminal control calling the transaction manager with a request to attach the task, and the subsequent trace entry showing the transaction manager calling dispatcher domain with a request to attach the task. The time stamps on the two trace entries tell you the time that elapsed between the two events. That is equal to the time taken for the task to be attached.

What to do next



dfhs1_using_trace.html | Timestamp icon Last updated: Monday, 5 June 2017