Question 440
The previous outage involved this problem: A Cloud SQL instance runs at very low utilization for six months after a major customer moves away. What long-term control addresses the risk?
Explanation
Correct Answer
Review sustained utilization after customer and workload migrations. The original failure arose from: The service size is not revisited after a sustained workload decline. The control reduces that failure mode.
Incorrect Answers
For the situation in which a Bigtable table consumes unexpectedly large storage because multiple historic cell versions are retained indefinitely, it makes sense to review Bigtable column-family retention policies as ingest formats evolve. Here, the service size is not revisited after a sustained workload decline, so the safeguard misses the recurring fault.
The stated control responds to a failure mode where a transactional workload uses a read replica for all reads, including read-after-write order confirmation, and customers see stale states. It does not prevent this one: the service size is not revisited after a sustained workload decline.
This would review row-key query shapes and timeout budgets before releasing new analytics access. It is useful when Bigtable read errors increase for one table while node CPU is not high, but it does not prevent the issue here: the service size is not revisited after a sustained workload decline.