CloudOps Engineer – Associate
AWS CloudOps Engineer – Associate (SOA-C03)
Independent preparation for AWS Certified CloudOps Engineer – Associate, formerly SysOps Administrator – Associate — the new blueprint, 57 study points across five note pages, and the operator instincts the exam rewards.
The operator’s exam. Where Solutions Architect asks what you would design, this one asks what you would do at 3 a.m. with an alarm firing. It validates deploying, managing and operating workloads on AWS: monitoring and remediation, reliability and business continuity, deployment automation, security controls, and networking operations — through the console, the CLI and infrastructure as code.
Renamed credential
AWS Certified SysOps Administrator – Associate became AWS Certified CloudOps Engineer – Associate with the SOA-C03 version. The new name applies only to people who pass the latest exam version; SOA-C02 passes keep the old title. The blueprint changed too — five domains with different weightings, and the earlier exam-lab section is gone.
Unofficial preparation
Independent study material, not affiliated with or endorsed by AWS, containing no official exam questions. Verify every figure on this page against the current SOA-C03 exam guide before booking.
The blueprint
| # | Domain | Weight | Where it is taught here |
|---|---|---|---|
| 1 | Monitoring, Logging, Analysis, Remediation, and Performance Optimization | 22% | Monitoring & observability |
| 2 | Reliability and Business Continuity | 22% | Availability, scaling & networking |
| 3 | Deployment, Provisioning, and Automation | 22% | Storage, deployment & provisioning |
| 4 | Security and Compliance | 16% | Governance & automation |
| 5 | Networking and Content Delivery | 18% | Availability, scaling & networking |
Three domains tied at 22% and nothing below 16% — the flattest associate blueprint AWS publishes. There is no domain you can trade away, and no single topic worth more than a fifth of the paper. Study breadth, then depth.
Cost and performance optimisation is no longer its own domain in SOA-C03, but cost work has not disappeared: it moved inside the monitoring and optimisation domain, which is why the cost management page stays in this course.
Exam mechanics
- 65 items, 130 minutes. 50 scored, 15 unscored pretest items mixed in unmarked.
- Scaled 100–1,000, pass at 720, compensatory across domains.
- Multiple choice and multiple response only — no exam labs in this version.
- Unanswered items score as incorrect and there is no guessing penalty.
- Expect console- and CLI-flavoured stems: specific alarm states, specific error text, specific metric names.
The five note pages
| # | Page | Study points | Covers |
|---|---|---|---|
| 1 | Monitoring & observability | 12 | CloudWatch metrics and custom metrics, alarms, Logs and Logs Insights, dashboards, anomaly detection, composite alarms |
| 2 | Availability, scaling & networking | 8 | Auto Scaling policies and lifecycle hooks, load balancing, multi-AZ patterns, VPC networking and security controls |
| 3 | Storage, deployment & provisioning | 17 | Storage and data management, deployment strategies, Lambda deployment configurations, advanced CloudFormation, Service Catalog |
| 4 | Cost management | 7 | Budgets and budget actions, Cost Explorer, Cost and Usage Reports, allocation tags |
| 5 | Governance & automation | 13 | AWS Config rules and remediation, advanced Systems Manager, Amazon Inspector, automation and governance patterns |
57 study points in total, weighted toward the deployment and governance material that the new blueprint pushed up.
The operator instincts the exam rewards
| Situation | What the exam wants |
|---|---|
| A metric you need is missing | Memory, disk and swap on EC2 are not default metrics — install the CloudWatch Agent |
An alarm reads INSUFFICIENT_DATA | Not a failure: no data points have arrived in the evaluation period yet |
| A fix has to happen without a human | Config rule detects, SSM Automation remediates — detection and remediation are two separate pieces |
| Instances need patching at scale | Systems Manager Patch Manager with patch groups by tag, not a login loop |
| Shell access must be auditable | Session Manager — no inbound ports, full session logging to CloudWatch Logs and S3 |
| An AMI must be rolled out to a running fleet | ASG Instance Refresh with a minimum healthy percentage, not manual replacement |
| Someone changed a stack by hand | CloudFormation drift detection, then bring the change back into the template |
| Spend must be capped, not just watched | Budgets with budget actions, not a Cost Explorer report nobody reads |
| A deployment must be reversible | Blue/green or canary with automatic rollback on alarm |
A three-week plan
Week 1 · Observability
Monitoring & observability twice. Build one alarm, one dashboard and one Logs Insights query in a real account — this domain is muscle memory.
Week 2 · Reliability and delivery
Availability, scaling & networking and Storage, deployment & provisioning. Deploy a stack, break it by hand, detect the drift.
Week 3 · Money and governance
Cost management and Governance & automation, then the practice exam timed.
Throughout · Live in the CLI
This exam assumes the CLI. Every console action you learn, do once more with aws — the stems are written by people who work that way.
Where to go next
- Compare all five tracks on the certification tracks page.
- Solutions Architect – Associate covers the design half of the same service surface and is the usual companion.
- Any service you cannot define belongs in the services glossary sweep.
Dernière mise à jour le 18 sept. 2026