In DBZ-9118, we introduced logic to validate the offset position; however, this change led to issues with Db2 where the validation may fail, not due to the position being invalid, but due to the fact that IBMSNAP_CAPMON is not always a reliable source for the validation.

The following holds true about IBMSNAP_CAPMON

• A new row is added when database CDC is disabled.
• A new row is added when the MONITOR_INTERVAL is reached, which is defined in the IBMSNAP_CAPPARAMS table.

I did look into IBMSNAP_PRUNCNTL but I found this table equally suffers from the same problem as IBMSNAP_CAPMON, where it was always empty.