Question

Your team runs a public-facing HTTPS API behind a global external load balancer. Users in Asia recently reported intermittent 'site down' complaints, but your existing latency dashboards showed no anomalies during those windows. You want a proactive, black-box signal that detects when the endpoint is unreachable from multiple geographic locations, and you want to be alerted only when the failure is confirmed from more than one region (to avoid noise from a single transient network blip). Which Cloud Monitoring configuration best meets this requirement?

Accepted Answer

Create a Cloud Monitoring uptime check against the public HTTPS endpoint with multiple selected checker regions, and configure an alerting policy that fires when the check fails from a minimum number of regions.

Answer

Add a log-based metric that counts HTTP 5xx entries in the load balancer logs and alert when the count exceeds zero in any one-minute window.

Answer

Instrument the application with Cloud Trace and create an alert on increased span latency for the API handler.

Answer

Enable Cloud Profiler on the backend service and alert when CPU utilization hotspots appear during the reported windows.