Create recovery points and synchronize log switches
A recovery point is a specific time that is valid for recovery of an object because no update activity was occurring for the object.
Standard recovery points are recorded in the IMS RECON data sets as image copy run time and database deallocation records. RMGR recognizes additional recovery points as times when no update activity for the object is recorded on the logs. RMGR also recognizes recovery points that are recorded in the RMGR repository by RMGR-simulated /DBD commands for DEDB areas and by the Log Synchronization (Log Sync) function.
The Log Sync function performs a switch of the logs for all IMS systems in an RMGR sharegroup and records this log switch activity in the repository as another type of recovery point. A Log Sync recovery point may be useful in a situation, such as disaster recovery, in which you need to coordinate a recovery across multiple IMS systems.
You can use the Find Recovery Points function to evaluate recovery points. If these recovery points do not occur frequently enough for you to meet service level agreements for application recovery, you can use the Hold Point of Consistency (HPC) and Log Sync function to create additional recovery points. You may also want to use the Log Analysis function to evaluate the online transaction update activity, BMP activity, and batch activity. Periods of inactivity are reported as recovery windows. Knowledge about whether you will find regular periods of inactivity may be useful in a recovery strategy.
For more information, see Create-recovery-points, Synchronize-log-switches, and Find-recovery-points-and-analyze-logs.
Related topic