Automatically locating checkpoints

You can automatically locate checkpoints to recover and reprocess messages in a nonshared-queues environment. The checkpoint locator applies to the RECOVERAB, RECOVERDM, and REPROCESS functions.

The RECOVERAB-DBRC function provides a list of logs, starting with the log that contains the first record of the checkpoint preceding the IMS crash (DUMPQ, PURGE, or SNAP). All relevant archived logs that follow are included in the list. IMS Queue Control Facility uses the list of logs to rebuild the queue.

The RECOVERDM-DBRC function provides a list of logs, starting with the log that contains the first record of the checkpoint preceding the DUMPQ or PURGE statement for normal IMS termination. All archived logs that contain these checkpoint records are included. IMS Queue Control Facility uses the list of logs to rebuild the queue.

The REPROCESS-DBRC function provides a list of logs, starting with the log that contains the checkpoint preceding the DUMPQ, PURGE, or SNAPQ statement set in the control cards. All relevant archived logs that follow are included. IMS Queue Control Facility uses the list of logs to rebuild the queue.

You can use either of the following two techniques to automatically locate checkpoints:

• Prepare JCL for the SELECT logic (and other steps in the function) by using the RECON data sets and skeletal JCL, and then submit this JCL.

  This technique uses a IMS Queue Control Facility batch logic.

• Make changes to the SELECT processing logic. Parameters are available in the CHKPT control card.

  In addition, add or modify DD statements in the SELECT step and submit the job.

  With this technique, the RECON data sets are read and the logs are dynamically allocated. The entire process completes in a single run.