Collecting data about an IMS dependent region wait or loop

Occasionally, there are problems in specific environments or certain problem types, that require special handling. If IMS can accept commands, you can set a trace.

1. If IMS can accept commands, use the following IMS command to set up the internal trace environment:

   /TRA SET ON TABLE nnnn

   where nnnn= can be DISP, SCHD, DLI, LOCK, or LATCH. Each must be entered separately.
2. Set the z/OS® system trace table size to 999K and turn on branch tracing with this command:

   TRACE ST,999K,BR=ON

3. If the problem is a wait, obtain two z/OS SVC dumps of the CTL, DL/I, suspicious dependent region, or CCTL, DBRC, and IRLM regions. If the problem is a loop, obtain two z/OS SVC dumps of the CTL, DL/I, suspicious dependent region, or CCTL, DBRC, and IRLM regions. Obtaining a second z/OS SVC dump will help differentiate normal IMS processing from the problem. Obtain a z/OS SVC dump with this series of commands:

   DUMP COMM=(dump title)
   R id JOBNAME=(j1,j2,j3,j4,j5),
   SDATA=(CSA,PSA,RGN,SQA,SUM,TRT),END

   j1

   is the IMS CTL or DBCTL region job name

   j2

   is the IMS DL/I region job name

   j3

   is the suspicious IMS dependent region job name, if any

   j4

   is the IRLM region job name (if IRLM DB locking is used)

   j5

   is the DBRC region job name

   In the previous example,
4. Reset the z/OS system trace table to its original settings.