Do I need to stop the PMA server and/or SIP server when running the PMA reorganization job APCBJREO?
APCBJREO is normally a long running job because of the large amount of records to remove from the KSDS files. This could cause some issues when there is set an ENQ against the whole file when doing the reorg, or even could cause a data integrity problem in the files.
The KSDSCST cluster access is being managed during the run of APCCASSB. As for the other VSAM clusters (job cluster, alert file cluster, etc), PMA calls a subroutine to manage IO on those, which has a ENQ/DEQ mechanism in place.