Forced recovery processing

For example:

Forced recovery processing provides an automated, integrity-assured facility for recovering databases following an Emergency Restart during which database allocation errors have occurred. These errors might have been caused by hardware failures (such as, an inoperative DASD controller or HDA) or by MVS problems (such as, a path to the DASD unit containing the affected databases has been varied offline). In either case, the restart completes successfully, but the databases that could not be allocated are marked as needing recovery processing and cannot be used by the IMS system. Recovery or Batch Backout processing of these databases is required before they can be /STA -rted and made available to the IMS system.

The typical procedure when this situation occurs is to shutdown the IMS system, correct the cause(s) of the allocation failures, set up and submit the required database recovery job streams and restart the IMS system.

Forced Recovery mode processing in RCF is a viable alternative for this scenario. To use it, simply perform the following:

• Shutdown the IMS system (/CHE FREEZE).
• Correct the causes of the allocation failures.
• Restart the IMS system with TRCF=F.

1. When the MTO specifies TRCF=F, RCF issues a prompting message to verify the request for Forced Recovery processing.

2. RCF checks to ensure that the previous shutdown of the IMS system was normal and that AUTO=N has been specified.

   If the previous shutdown of the IMS system was not normal or the RDS is unusable, RCF processing mode is reset to the normal TRCF=Y setting. When AUTO=Y is detected, the MTO is prompted for permission to change it to AUTO=N. If this is allowed, the change is made and Forced Recovery processing continues; if not, the processing mode reverts to TRCF=Y.

3. RCF prompts the MTO to specify the checkpoint-ID to be used for database recovery processing.

   The value specified in the MTOs response must be a valid system checkpoint from the IMS system session prior to the /ERE attempt. This can be obtained from the DFS994I messages on the IMS system Hardcopy Log. It must be a valid checkpoint for this IMS system, and must be prior to the Current and Fast Path checkpoint-IDs in the RDS data set.

4. Analysis of, and any required actions relating to, MSDB integrity are performed in the normal fashion.

5. If Forced Recovery mode is still in effect, the BMC Fast Path Recovery Utility (FRU) product is invoked using the checkpoint information specified by the MTO.

   The utility performs DEDB recovery and gathers information needed for creating and submitting any required Batch Backout job streams.

6. The MTO is prompted to enter the IMS restart command to be passed to the Control Region.

   The permissible responses are either a warm start or a cold start.