Question 530
During an operational security assessment at a public-sector technology agency, the team documents this issue: A partial rollout requires quick reversal but the pipeline stores no prior healthy artifact. The affected delivery process supports citizen applications and reusable infrastructure; accountability and service availability depend on controlled delivery. Which TWO responses form a sound implementation-and-verification pair for this issue? (Select TWO.)
Explanation
Correct Answers
Correct corrective action. Deploy to a small percentage of production traffic, monitor meaningful error and latency signals, and automatically stop or roll back when thresholds fail. The action targets the recorded failure (A partial rollout requires quick reversal but the pipeline stores no prior healthy artifact), and is intended to reduce the corresponding risk: Instant full traffic exposure can turn a contained defect into a service-wide outage.
Correct verification. A deliberately failing canary exceeds its threshold, stops traffic expansion and restores the last healthy release. The measurement or controlled test provides evidence for the issue (A partial rollout requires quick reversal but the pipeline stores no prior healthy artifact) and complements the selected corrective action rather than simply repeating a recommendation.
Incorrect Answers
The same deployment task can be retried after a simulated interruption without duplicate schema or business effects. It is not the complementary risk control for canary rollout with metric-based rollback.
Apply backward-compatible database migrations before code that depends on them and remove deprecated columns only after older clients retire. The paired controls concern canary rollout with metric-based rollback, not that measure.
Provide a narrow, reviewed hotfix path that validates the corrective change and promotes a traceable artifact while preserving production safeguards. The two-part canary rollout with metric-based rollback control remains unsatisfied.