Google Cloud Professional DevOps Engineer Intermediate Practice Exam: Medium Difficulty 2025

Your organization recently migrated to Google Cloud and created several projects manually. You need to standardize project creation so every new project automatically has: a shared VPC attachment, required APIs enabled, centralized logging sinks, and a baseline set of IAM bindings. You also need an auditable, repeatable process that different teams can use with minimal permissions. What should you do?

Your security team requires that production projects cannot have owner-level access granted to human users, and that IAM changes must be centrally controlled. However, developers still need to deploy to production using CI/CD. Which approach best meets these requirements?

A team uses Cloud Build to deploy a container to Cloud Run. They want to promote the same artifact through dev, staging, and prod, and require a manual approval gate before production. They also want the pipeline to be traceable back to a Git commit and to avoid rebuilding the image for each environment. What should they do?

You manage a microservices platform on GKE. You want to reduce deployment risk by using canary releases, gradually shifting traffic while monitoring key metrics. If error rates rise above a threshold, the rollout should automatically stop and roll back. Which solution best fits?

A CI pipeline builds containers and deploys to multiple projects. Security requires that build steps cannot exfiltrate secrets and that access to deploy credentials is tightly scoped. The team currently uses long-lived service account keys stored in the repo. What should you do?

Your service has an SLO of 99.9% availability per month. In the last week, several incidents consumed most of the error budget. Product wants to launch a risky new feature immediately. As the DevOps lead, what should you do?

An on-call team is overwhelmed by alerts from a GKE-based service. Many alerts are not actionable and fire repeatedly during known maintenance events. You need to improve signal-to-noise while still detecting real user-impacting incidents quickly. What should you do?

A backend service on Cloud Run experiences occasional latency spikes. You need to identify whether the spikes correlate with specific upstream dependencies and deployments. You also want to drill into individual requests across services. What is the best approach?

You operate multiple services across several projects. You need a centralized view of health: dashboards, alerting policies, and SLOs that aggregate across projects and environments. You also need teams to retain control of their own project resources while enabling central monitoring. What should you do?

A GKE-hosted API has strict latency targets. Load increases during peak hours, and the service becomes CPU-bound, causing higher p95 latency. You want to maintain latency SLOs cost-effectively without overprovisioning. What should you do?