Features and Benefits
IMS Program Restart Facility (PRF) automates the restart of abended Batch Message Processing (BMP) regions and DL/I batch jobs. Restarting IMS jobs is potentially error prone, as manual JCL changes are required to add the checkpoint ID and the log data set containing the restart checkpoint. If either of these manual changes is incorrect or omitted, or if a job that should be restarted is not restarted, database corruption can result.
IMS Program Restart Facility ensures that:
A batch job that should be restarted is forced to restart.
The proper log data set and the proper checkpoint ID are used for restart.
In a data sharing IMSPlex environment, PRF provides the additional feature of allowing a BMP to be restarted using a different IMS system. Taking advantage of this feature provides improved restart times in the event of a system outage.
IMS Program Restart Facility is easily implemented, and typically requires no application changes. While an IMS batch job is running, PRF intercepts and writes checkpoint records to two small data sets. These data sets are deleted when the batch job completes successfully. In the event of an abend, however, the data sets remain in place, and are used to restart the job. Since PRF uses the small DASD data sets, tape mounts for IMS logs are avoided.
IMS Program Restart Facility Version 2 provides the following enhancements:
Bypass Checkpoint Processing, which throttles the frequency of application checkpoints. This can reduce logging and improve batch job elapsed times.
Flag non-zero return code as an abend, which causes PRF to treat a batch job with a non-zero return code as an abend, ensuring that the job is restarted properly.
Provide an option to hold and restart a BMP through operator commands – without actually stopping the job.
These features are controlled through an ISPF interface, which allows specification options by jobname, PSB name, program name, or IMSID.
