For anyone wondering about the purpose of the R19 CDBAPARM(UTIL) parm "SYSCOPY-ROWS".
Choice is ‘1’ or ‘2’, with default=2
Explanation is that an ICTYPE=S row written by (PFL) LOAD REPLACE SET-COPYPENDING NO should result in 2 rows written to SYSCOPY.
These 2 SYSCOPY rows were introduced to support PFL customers who run the following sequence on a tablespace
- CA Fast Load (PFL) LOAD LOG(NO) NOCOPYPEND
- This breaks the recovery chain. Tablespace is changed, but the changes are not logged + copy pending is suppressed
- CA Quick Copy (PQC) COPY SHRLEVEL CHANGE
- This restores recoverability
If COPY immediately follows LOAD, then the only valid START_RBA for the copy PQC can find is the START_RBA of the previous ICTYPE=’S’ row. But that is guaranteed only if the ICTYPE=’S’ row was inserted by CA Fast Load.
Since SYSIBM.SYSCOPY doesn’t allow to specify vendor of the utility, CA Fast Load can be set to insert two almost identical ICTYPE=’S’ rows. CA Quick Copy can find these rows and determine that the ICTYPE=’S’ row was inserted by CA Fast Load (and not by IBM LOAD).
[Answer kindly provided by Michael Kalous]