ACE Domain 3: Ensuring the successful operation of a cloud solution - Complete Study Guide 2026

What Domain 3 Actually Covers

Domain 3, "Ensuring the successful operation of a cloud solution," is the operations-heavy section of the Associate Cloud Engineer exam. Where Domain 2 is about building a solution, Domain 3 is about keeping it running: watching dashboards, reading logs, managing running compute and storage resources, and rolling out changes without breaking production. If you have ever been paged at 2 a.m. for a full disk or a stuck deployment, this domain will feel familiar.

Google's exam guide groups Domain 3 tasks around four practical clusters: managing compute resources, managing storage and database solutions, managing networking resources, and monitoring/logging solutions. A newer thread running through all four is process automation and observability tooling, including AI-assisted operations. For the full breakdown of how this domain fits alongside the other three, see the complete guide to all four ACE content areas.

Monitoring, Logging, and Observability

This is the single densest topic inside Domain 3. You need working familiarity with:

• Cloud Monitoring: creating dashboards, uptime checks, alerting policies, and notification channels (email, SMS, Pub/Sub, PagerDuty-style integrations).
• Cloud Logging: log routing with sinks, log-based metrics, exclusion filters, and viewing logs across projects with the Logs Explorer.
• Error Reporting and Cloud Trace/Profiler: recognizing when each tool is the right answer for latency versus error-rate versus resource-consumption questions.
• Audit logs: distinguishing Admin Activity, Data Access, System Event, and Policy Denied logs, and knowing which are on by default.

Expect scenario questions rather than pure definition questions. A typical prompt might describe a spike in 5xx errors on a Compute Engine-backed load balancer and ask which combination of Cloud Monitoring and Cloud Logging configuration will notify the team fastest with the least noise.

Managing Compute Resources

Domain 3 assumes you already deployed compute resources (that was Domain 2's job) and now asks you to keep them healthy. Core skills include:

• Managing instance groups: resizing, updating instance templates, and rolling updates versus rolling restarts.
• Adjusting autoscaling policies for Compute Engine managed instance groups and understanding cooldown periods.
• Managing GKE clusters day-to-day: node pool upgrades, workload scaling, and reading `kubectl` output for troubleshooting.
• Understanding Cloud Run and Cloud Run functions revisions, traffic splitting, and rollback behavior when a new revision misbehaves.
• SSH access, OS Login, and startup/shutdown script troubleshooting on Compute Engine.

Managing Storage and Database Solutions

Operational storage questions focus less on choosing a storage product (that's planning) and more on keeping existing storage healthy and cost-controlled:

• Cloud Storage lifecycle management rules, retention policies, and object versioning cleanup.
• Monitoring Cloud SQL instances: read replica lag, storage autogrowth, and backup/restore procedures.
• Managing Cloud SQL and Spanner maintenance windows without causing unplanned downtime.
• Bigtable and Firestore capacity and performance monitoring basics.
• Import/export and migration jobs using `gcloud` and the console, including troubleshooting failed transfers.

Managing Networking Resources

Networking in Domain 3 is about maintaining connectivity and diagnosing breakage, not designing VPCs from scratch:

• Troubleshooting firewall rules that block expected traffic, including priority and direction conflicts.
• Verifying VPC peering, Cloud NAT, and Cloud VPN/Interconnect connectivity issues.
• Using Network Intelligence Center tools (Connectivity Tests, Firewall Insights) to diagnose problems.
• Managing DNS records in Cloud DNS and diagnosing propagation or misconfiguration issues.
• Reviewing load balancer health checks when backends are marked unhealthy.

This overlaps with the access-control concepts tested in Domain 4, so review both together rather than in isolation - a firewall troubleshooting question can straddle both domains.

Deployment and Process Automation

Domain 3 also covers keeping deployments repeatable and safe:

• Deployment Manager and Terraform basics: applying, updating, and rolling back infrastructure changes.
• Cloud Build triggers, build steps, and troubleshooting a failed CI/CD pipeline.
• Application Design Center's role in generating and maintaining deployable application blueprints.
• Using `gcloud` scripting and Cloud Scheduler for recurring operational tasks like backups or cleanup jobs.

AI-Assisted Operations with Gemini Cloud Assist

The current exam guide incorporates Gemini Cloud Assist and Gemini CLI as operational aids, alongside Agent Runtime on the Gemini Enterprise Agent Platform and Google Antigravity. For Domain 3 purposes, you don't need deep AI engineering knowledge - you need to recognize where these tools fit into day-to-day operations:

• Gemini Cloud Assist can summarize error patterns from Cloud Logging and suggest remediation steps inside the console.
• Gemini CLI can help generate or explain `gcloud` commands during troubleshooting sessions.
• These tools assist human operators; they don't replace the need to understand what Cloud Monitoring, Logging, and IAM are actually doing under the hood.

Scheduling Domain 3 Into Your Prep

Domain 3 rewards hands-on repetition more than reading. If you're following a multi-week plan like the one in the ACE study guide for passing on the first attempt, place Domain 3 after you've already provisioned resources in Domain 1 and 2 labs, so you have something real to monitor and troubleshoot.

Two labs beat ten flashcards for this domain. If your time is limited, prioritize Cloud Monitoring and Cloud Logging labs above everything else in Domain 3.

Registration, Format, and Timing

Domain 3 questions are mixed throughout the exam rather than grouped in a section, so mechanics matter as much as content knowledge. The ACE exam runs 2 hours with 50-60 multiple-choice and multiple-select questions, delivered online-proctored or at a Pearson VUE test center. Registration happens through CM Connect/CertMetrics, and the standard exam fee is $125 USD plus tax. For a full cost breakdown including the renewal path, see the ACE certification cost guide.

There are no prerequisites, but Google recommends 6+ months of hands-on Google Cloud experience - and Domain 3's scenario-based troubleshooting questions are exactly where that experience shows. Candidates get up to 4 attempts in a 2-year period, with waiting periods between failed attempts, so it's worth building real operational familiarity rather than rushing a retake. If you're weighing how difficult this section is compared to the rest of the exam, the ACE difficulty guide and pass rate analysis both discuss where candidates commonly lose points.

Before exam day, run a few timed practice sets on the ACE practice test platform so you get comfortable pacing through mixed-domain scenario questions under the 2-hour clock. Repeating troubleshooting-style questions on a full-length practice exam is one of the fastest ways to convert lab experience into exam-ready recall.

Frequently Asked Questions