Google Cloud Professional Cloud DevOps Engineer

About the Professional Cloud DevOps Engineer exam

Google Cloud Professional Cloud DevOps Engineer is for people who run production systems on Google Cloud and the pipelines that feed them. It checks whether you can set up an organization so teams can ship safely, build CI/CD that is fast and trustworthy, apply site reliability engineering to decide how much risk a service can take, and find out what went wrong when it does. Google recommends three or more years of industry experience, including at least a year managing production systems on Google Cloud. There is no prerequisite.

Questions describe a team with a problem: an alert that pages too often, a rollout that broke half the users, a logging bill that doubled, a build that pulls an unscanned image. You choose the fix Google's tooling and SRE practice would recommend. Wrong answers often work but add manual toil, alert on causes instead of symptoms, or trade away reliability the error budget doesn't allow.

Exam format

• Questions: 50 to 60, multiple choice and multiple select.
• Time: 2 hours.
• Passing score: Google reports pass or fail and does not publish a cut score.
• Cost: 200 USD plus tax, online with a remote proctor or at a Pearson VUE test center. English and Japanese.

What the exam covers

The official exam guide has five sections:

• Bootstrapping and maintaining a Google Cloud organization (~20%). Resource hierarchy, Shared VPC, service accounts and org policies; infrastructure as code with Terraform, Infrastructure Manager and Config Connector; GKE fleets; Cloud Workstations; and AI help from Gemini Code Assist and Gemini Cloud Assist.
• Building and implementing CI/CD pipelines (~25%). Cloud Build, Cloud Deploy and Artifact Registry; canary, blue/green and rolling releases; Secret Manager, Parameter Manager and Cloud KMS; and supply chain security with Artifact Analysis, Binary Authorization and SLSA.
• Applying site reliability engineering practices (~18%). SLIs, SLOs and SLAs, error budgets, capacity planning with quotas and reservations, autoscaling, and limiting the damage of an incident by draining traffic, adding capacity or rolling back.
• Implementing observability practices and troubleshooting issues (~25%). Ops Agent and OpenTelemetry, log-based and custom metrics, Managed Service for Prometheus, synthetic monitors, log routing and redaction, alerting policies, and reading distributed traces.
• Optimizing performance and cost (~12%). Active Assist recommendations, Spot VMs, committed-use discounts, the cost of observability itself, and per-workload savings on GKE, Cloud Run and Compute Engine.

How the practice questions here are organised

There is one 10-question practice test. It leans on SRE and observability, which together make up over 40% of the exam: error budget policies, incident command roles, toil, burn-rate alerts on SLOs, log-based distribution metrics and linking traces to logs. It also covers Binary Authorization attestations, a Cloud Deploy canary promotion, secure development environments, and cutting logging ingestion cost with exclusions. Every question explains the right answer and each wrong one, with a link to the documentation behind it.

There is also a full practice exam that copies the real one: 55 questions in 120 minutes, weighted across the five sections as the guide does, with answers hidden until you submit. Google doesn't publish a cut score, so the mock passes you at 70%.

A study plan that works

• Take the practice test cold, then read every explanation, including the ones for answers you guessed.
• Read the SRE book chapters on SLOs, error budgets, toil and incident management. A large part of this exam is applying those ideas, not memorising product settings.
• In a test project, wire up Cloud Build to Artifact Registry to Cloud Deploy, then add a Binary Authorization policy. Seeing a blocked deployment fixes the concepts in place.
• Create an SLO in Cloud Monitoring and a burn-rate alert on it, and route some logs to BigQuery with an exclusion filter.
• When most practice questions feel routine, sit the full practice exam in one go, timed. Treat a score in the mid 70s or above as a sign you're ready to book.

These questions are original, written from the public Google Cloud documentation and the official exam guide. They are not real exam questions, and this site is not affiliated with Google.