Recovering both IMS and Db2 subsystems at the remote site
You may run the IMS and Db2® combined recovery jobs at the remote site.
Procedure
- Ensure that you have restored the recovery PDS.
- Locate the Db2 and IMS jobs
that execute the BSY#CDR or ARY#CDR program. Execute these jobs to
determine the most current timestamp that can be used for recovery,
and to update the jobs in the recovery PDS jobs with the correct timestamp.
These jobs will run the coordinated disaster recovery timestamp determination process, and will then update the related recovery PDS jobs with that timestamp. The Db2 job might also have additional steps to reconstruct the BSDS data sets.
These jobs compare the highest available archive timestamp of IMS and Db2. The lowest of these timestamps can be used for coordinated recovery. The Db2 job updates the Db2-ssidCRCR member with this timestamp. The IMS job updates all application recovery jobs.
If the Db2 subsystem is a non data-sharing subsystem, then RBAs are used instead of LRSN; if this happens, then Db2 will have to identify an earlier point in time that can be matched with an RBA.
After a recovery timestamp is determined, IMS Recovery Expert must ensure that the SLB that is used for recovery is earlier than this timestamp. If it is not, an older backup will be used, and IMS and Db2 SLB recovery jobs will be modified to reflect that.
- Follow steps 4 through 5 of Recovering your system at the remote site to perform the IMS system point in
time recovery. Consult the IBM®
Db2 Recovery Expert for z/OS® User's Guide for the recovery steps
for Db2.
Note: If IMS Recovery Expert and Db2 Recovery Expert are sharing the same product repository, then you must consider the following points:
- IMS and Db2 both have jobs that delete, redefine, and reload the repository data sets. You only have to run one of these jobs. Run the job that was generated later to ensure that the repositories contain data for both IMS and Db2.
- It is recommended that you take a SLB of your subsystems after full recovery.