Question 880
A postincident review documents: A production Cloud SQL instance is created with the correct size but monitoring shows no CPU or storage data in the operational dashboard. Which ongoing check targets the underlying failure?
Explanation
Correct Answer
Include telemetry and alert delivery in database acceptance tests. This is preventive rather than a one-time repair; it addresses the underlying failure: Required metrics collection or dashboard resource filters were not configured for the new instance.
Incorrect Answers
This would use compatibility gates for database DDL changes in continuous delivery. It is useful when an automated database schema deployment runs before a new application release but introduces an incompatible column change, but it does not prevent the issue here: required metrics collection or dashboard resource filters were not configured for the new instance.
For the situation in which a Bigtable table is created with dozens of column families for each sensor field and maintenance becomes difficult, it makes sense to review family count and GC purpose before deploying Bigtable schemas. Here, required metrics collection or dashboard resource filters were not configured for the new instance, so the safeguard misses the recurring fault.
This control would schedule failover validation after infrastructure or database driver changes; it responds to the risk in which a database service looks healthy after deployment yet its automatic failover path has never been exercised with the application. It would not address the current root cause: required metrics collection or dashboard resource filters were not configured for the new instance.