You need an email notification when a Compute Engine instance exceeds 90% CPU for more than 15 minutes using Google Cloud services. What should you do?

The correct approach is to use Cloud Monitoring's built-in alerting capabilities. Cloud Monitoring directly integrates with Compute Engine metrics, allowing you to create an alerting policy based on CPU utilization. You can define a threshold (90%) and a duration (15 minutes) and then configure email as a notification channel. This is the most direct, managed, and efficient way to achieve the desired outcome within Google Cloud. Creating a consumer Gmail account and a custom script is overly complex and unmanaged, requiring external infrastructure and maintenance. Creating a Cloud Monitoring workspace and a script for custom metrics is unnecessary because CPU usage is a standard metric already collected by Cloud Monitoring. Creating a logs-based metric is also incorrect; CPU usage is a metric, not a log entry, and Cloud Logging is not the primary service for real-time metric-based alerting.