Recovery processing

The first request for I/O to the control data set, following a multi-record update failure, detects the failure, and recovery processing begins. Other attempts to access the control data set are queued behind the current request until either manual recovery is required or automatic recovery is successful.

When a multi-record update failure is first detected, DFSMSrmm attempts automatic recovery processing which requires no operator intervention. If a journal data set matching the corrupt control data set is found, DFSMSrmm issues message EDG2111I to notify the operator that automatic recovery processing is starting. DFSMSrmm issues message EDG2115I if automatic recovery is not possible and manual recovery is required. Manual recovery might be needed under these conditions:

The journal data set is not defined in the initialization parameters
The journal and control data sets do not match
The journal data set has been disabled in response to message EDG2103D
The journal data set update has been ignored in response to message EDG2103D

Manual recovery requires operator intervention. See z/OS DFSMSrmm Managing and Using Removable Media for operator procedures for responding to messages DFSMSrmm issues during recovery processing.

Recommendation: Use EDGHSKP with the CATSYNCH parameter to synchronize catalogs after recovery.