When a CA DISK job is running to backup VSAM RLS datasets, it can happen that CICS transactions trying to access these VSAM datasets get AFCK abend.
How is it possible to bypass this abend?
Component : CA Disk Backup and Restore
From the IBM documentation at the following link:
The transaction issued a file update request (READ UPDATE, WRITE or DELETE) against an RLS mode data set for which a DFSMSdss non-BWO backup was in progress.
The transaction is abnormally terminated with a CICS transaction dump. CICS processing continues.
All new file update requests are prohibited when a non-BWO backup is in progress for an RLS mode data set. This restriction is automatically lifted when the backup completes. (A non-BWO backup is any type of backup operation other than a Backup While Open backup.) When the backup has completed, retry the transaction.
Based on that, to possibly prevent these abends, it is suggested to use the CA DISK BACKUPCC command for the VSAM RLS datasets backups. With BACKUPCC, updates will still be possible even while the backup is currently running.
The following link will go in details about BACKUPCC :