A production Cloud Run service is returning HTTP 500s. You want to be alerted when errors exceed 15% of requests during a time window. What should you do?

The correct option is to create an alerting policy in Cloud Monitoring. Cloud Monitoring is designed for exactly this purpose: to collect metrics, monitor resource health, and trigger alerts based on predefined conditions. An alerting policy can directly monitor Cloud Run service metrics, such as request counts and error rates, and send notifications when the error percentage exceeds 15% within a specified time window. The other options are less efficient or incorrect. Manually refreshing the Cloud Run page daily is not an automated alerting solution. Using Cloud Functions with Cloud Monitoring API calls, triggered by Cloud Scheduler or Cloud Composer, is overly complex and unnecessary. Cloud Monitoring provides native alerting capabilities that are simpler to configure and more robust for this use case. Cloud Composer is also generally used for orchestrating complex workflows, not simple metric-based alerting.