Managing application recovery checkpoint restart

IMS Recovery Expert provides the capability to perform checkpoint restart of failed application recovery jobs. This is an optional feature that must be activated by allocating the BSYCHKPT repository and updating the BSYV220 clist to specify the data set in the RBRCHKPT variable.

For more information about this, see step 3. If this repository is set up, when the application recovery JCL is created, the BSYCHKPT DD statement is included in the job, and that activates checkpoint restart. No additional tasks are required to activate this feature. When the application recovery job executes for the first time, information is recorded in this repository that can be used for subsequent executions if there are any errors.

If the application recovery job completes successfully, the information in the BSYCHKPT repository is deleted. If the same application recovery job is executed again, then it will attempt to perform the entire recovery process again. If any errors occur while performing any of the steps or jobs required to complete the application recovery for any databases, then the information remains unchanged in the repository. If the job is resubmitted, then IMS Recovery Expert will use the information in the BSYCHKPT repository as well as information in the RECON data sets to determine what processes failed on the prior execution and need to be rerun. After all processes complete for all databases, the information in the repository is deleted.

Note: If a failure occurs in an application recovery job and checkpoint restart is active, when the job is resubmitted, the execution environment is verified to ensure that changes were not made. You cannot change the type of recovery being performed with checkpoint restart.