Google Cloud DevOps Certification Practice Test

To ensure alerts are triggered only for consistent high utilization over a specified period, setting up an alert with a mean threshold and a rolling window is key. The option that involves creating an alert from a log-based metric with a 5-minute rolling window and a mean threshold of 80% aligns perfectly with this need.

This setup means that the alert will monitor the average utilization over a rolling period of 5 minutes. By specifying a mean threshold of 80%, you require that, on average, the utilization remains above this level for the entire rolling window. This approach efficiently filters out short-lived spikes in utilization, ensuring that the alerts you receive signify a sustained condition rather than fleeting instances.

In contrast, other options might not effectively limit alerts to consistent usage patterns. For example, using a mean threshold of 90% could lead to fewer alerts, as it would only consider a higher average, possibly missing critical incidents with high but not extreme utilization. Similarly, directing notifications to a Cloud Run application for filtering, while a valid method for complex alerts, may not natively handle the rolling window mechanism without additional logic, making it less straightforward for the requirement presented. Additionally, configuring alerts with just a metric absence condition would not necessarily ensure that the alerts reflect sustained