AWS SOA-C03 FAQ for exam format, topics, prep strategy, practice, and common candidate traps.
SOA-C03 is AWS’s associate-level CloudOps certification. It tests whether you can operate workloads safely after deployment: monitor, troubleshoot, remediate, secure, restore, and keep services available under operational pressure.
According to the current AWS guide, the target candidate should have about 1 year of AWS experience across deployment, management, troubleshooting, networking, and security, plus related operations experience.
As of April 11, 2026, AWS still lists:
| Exam fact | Current official value |
|---|---|
| Duration | 130 minutes |
| Questions | 65 |
| Response types | Multiple choice and multiple response |
| Passing score | 720 scaled score |
AWS uses a compensatory scoring model. You do not need to pass every domain separately; you need to pass the exam overall.
No. The current AWS exam guide lists only multiple-choice and multiple-response question types.
That does not mean hands-on work is optional. It means the hands-on requirement shows up indirectly: AWS expects you to recognize the right safe action, not just repeat service names.
The current guide uses five weighted domains:
| Domain | Weight |
|---|---|
| Monitoring, Logging, Analysis, Remediation, and Performance Optimization | 22% |
| Reliability and Business Continuity | 22% |
| Deployment, Provisioning, and Automation | 22% |
| Security and Compliance | 16% |
| Networking and Content Delivery | 18% |
That means the exam is heavily operational. Observability, continuity, provisioning, and networking all matter more than abstract architecture discussion.
They are hard in different ways:
| Exam | What it rewards most |
|---|---|
SAA-C03 |
Architecture design trade-offs and service-fit decisions |
SOA-C03 |
Operational judgment, remediation order, continuity, and troubleshooting |
If you are stronger in day-two operations, incident handling, monitoring, backups, routing, and IAM troubleshooting, SOA-C03 usually feels more natural than SAA-C03.
The strongest baseline is:
You do not need deep software engineering depth, but you do need to reason clearly about operational consequences.
If time is limited, bias toward the places candidates usually lose easy points:
RTO, and RPO reasoningSOA-C03 is not mainly testing:
AWS can still mention those things in a stem, but the scored decision usually lands in the operations lane: monitoring, change control, recovery, access, or network diagnosis.
If you want the smallest hands-on loop that still changes how you answer questions, do this:
RTO or RPO.That is enough to stop answering from buzzwords alone.
| Weak area | Why people miss it |
|---|---|
| alarm vs remediation vs recovery | candidates blur detection with the corrective action |
| Multi-AZ vs backup vs DR | these sound similar but solve different operational problems |
| EventBridge vs Systems Manager vs Lambda | routing, runbook execution, and code execution are different lanes |
| Access Analyzer vs CloudTrail vs policy simulator | audit, history, and policy evaluation are not the same thing |
| CloudFront vs Route 53 vs Global Accelerator | content delivery, DNS routing, and global traffic handling blur together |
Use a narrow loop:
In the last week, stop trying to learn AWS from scratch. Compress and stabilize:
2 timed mixed sets.If you still confuse backup vs replication, alarm vs remediation, or Route 53 vs CloudFront vs Global Accelerator, keep drilling. Those are expensive exam misses.
Trust the live AWS docs in Resources first:
Those are more reliable than older blog posts, course thumbnails, or recycled third-party objective lists.