Question 620
An API is intermittently slow but its CPU usage is low, so managers conclude it is healthy. The service is in its initial cost-and-reliability governance cycle, with agreed baselines required. What should the team verify FIRST before committing to a solution?
Explanation
Correct Answer
The required decision is review Cloud Monitoring latency distribution and correlated traces over affected periods. It addresses this operating condition: an API is intermittently slow but its CPU usage is low, so managers conclude it is healthy.
Incorrect Answers
“Measure historical user-facing availability and define tolerable failure budgets” concerns service objective definition. It does not address the stated condition (an API is intermittently slow but its CPU usage is low, so managers conclude it is healthy. The service is in its initial cost-and-reliability governance cycle, with agreed baselines required); the closer match is review Cloud Monitoring latency distribution and correlated traces over affected periods.
“Map single-zone dependencies and actual backup restoration steps” concerns cross-zone resilience drills. It does not address the stated condition (an API is intermittently slow but its CPU usage is low, so managers conclude it is healthy. The service is in its initial cost-and-reliability governance cycle, with agreed baselines required); the closer match is review Cloud Monitoring latency distribution and correlated traces over affected periods.
“Inspect correlated Cloud Logging entries and Cloud Trace spans for failed requests” concerns log-based incident investigation. It does not address the stated condition (an API is intermittently slow but its CPU usage is low, so managers conclude it is healthy. The service is in its initial cost-and-reliability governance cycle, with agreed baselines required); the closer match is review Cloud Monitoring latency distribution and correlated traces over affected periods.