Question 720
After an incident where two systems pass primary-key row comparisons but downstream reports disagree because CDC excludes one business-critical table, which recurring control would best prevent the same defect?
Explanation
Correct Answer
Reconcile CDC table inventories with downstream consumer requirements. This is preventive rather than a one-time repair; it addresses the underlying failure: Replication scope was defined without a complete list of application and reporting dependencies.
Incorrect Answers
This would verify identity and sequence state after every relational data load. It is useful when an Oracle application's sequence-generated invoice IDs collide with target rows after importing data into PostgreSQL, but it does not prevent the issue here: replication scope was defined without a complete list of application and reporting dependencies.
The proposed check would validate insert, update and delete event handling separately, helping with cases where a continuous replication stream reports success but deletes on the source remain visible at the target. The recurring exposure described here is that replication scope was defined without a complete list of application and reporting dependencies.
The failure in the question arises because replication scope was defined without a complete list of application and reporting dependencies. A safeguard to monitor CDC offsets and verify completeness after restarts serves situations where a change-capture pipeline restarts after a network outage and silently skips a period of source commits rather than this one.