C90.02 Exam Guide: Identify the AWS Exam, Prepare for SOA-C03, and Schedule with Confidence
“C90.02” does not match an exact AWS certification identifier in the supplied official material. The closest official identifiers are SOA-C02, the former AWS Certified SysOps Administrator – Associate exam, and SOA-C03, the updated AWS Certified CloudOps Engineer – Associate exam. SOA-C03 validates the ability to deploy, manage, and operate workloads on AWS. This guide helps you determine which exam your catalogue entry means, avoid preparing from an obsolete blueprint, assess your readiness, and choose the next practical step before booking.
What does C90.02 refer to?
Treat C90.02 as a catalogue label that requires verification, not as a confirmed AWS exam code. AWS’s supplied comparison page identifies SOA-C02 and SOA-C03, while the official SOA-C03 exam guide describes SOA-C03 as AWS Certified CloudOps Engineer – Associate. Before purchasing preparation material or scheduling, compare the exam name and identifier in your AWS Certification account with the booking information on the official AWS page.
The most likely interpretation is that C90.02 points to the AWS operations certification family represented by SOA-C02 or SOA-C03. AWS states that SOA-C02 was the previous AWS Certified SysOps Administrator – Associate exam and that SOA-C03 is its updated version. AWS also identifies SOA-C03 as the AWS Certified CloudOps Engineer – Associate exam, formerly known as AWS Certified SysOps Administrator – Associate.
This distinction affects the entire study plan. SOA-C02 had six scored domains, including a separate Cost and Performance Optimization domain. SOA-C03 has five scored domains and gives greater weight to several operational areas. A resource labelled for the older exam may still explain useful AWS services, but it should not be treated as a current blueprint without checking its domain mapping and task coverage.
The safest identification check
Use the official AWS certification page to confirm the exam title, identifier, delivery options, languages, and current scheduling information. Then use the AWS exam guide to inspect the target candidate profile, content domains, task statements, in-scope services, and out-of-scope activities. If your provider continues to display only C90.02, ask it to state whether the entry maps to SOA-C02 or SOA-C03 before you commit study time.
What does the current exam validate?
SOA-C03 validates practical ability to deploy, manage, and operate workloads on AWS. It is aimed at CloudOps engineers and covers operational work rather than software development or the design of every component in a distributed architecture. Your preparation should therefore emphasize diagnosis, configuration, controlled change, security controls, resilience, and day-to-day use of AWS operations tools.
The operational tasks behind the exam
The official exam guide says the exam validates support and maintenance of AWS workloads according to the AWS Well-Architected Framework, operations through the AWS Management Console and AWS CLI, security controls for compliance, monitoring and troubleshooting, networking concepts, high availability and capacity requirements, business continuity and disaster recovery procedures, and incident identification, classification, and remediation.
These statements point to a decision-oriented exam. You need to recognize which operational action fits a symptom, constraint, or requirement. For example, a scenario may require you to distinguish an observability problem from an access-control problem, or a recovery requirement from a routine deployment task. Memorizing isolated service descriptions is less useful than learning how the services work together in an operational sequence.
AWS lists general knowledge expectations that include monitoring, logging, and troubleshooting; DNS, TCP, IP, and firewalls; high availability, performance, and capacity; one scripting language; one major operating system; cloud computing; containerization and orchestration basics; CI/CD and Git. These are preparation prerequisites in the target profile, not a promise that every candidate must hold a particular external credential.
What the exam does not ask you to become
The official guide places several job tasks out of scope, including designing distributed architectures, designing CI/CD pipelines, designing hybrid and multi-VPC networking, developing software, defining security, compliance, and governance requirements, developing ransomware defense strategies, assessing and planning resource capacity, analyzing costs and total cost of ownership, and managing business continuity. These boundaries help prevent over-study.
You still need enough understanding to operate services within those contexts. The practical distinction is between implementing or troubleshooting an existing requirement and designing the organization’s complete strategy. Study how to apply documented controls and operational procedures; do not turn the plan into an architecture-design or application-development curriculum.
Who should use this preparation plan?
The intended candidate has approximately one year of experience deploying, managing, troubleshooting, networking, and securing AWS workloads, plus at least one year in a related operations role such as system administrator. If you have less hands-on exposure, allow additional time to build fundamentals before relying on practice questions. If you already operate AWS workloads, use the blueprint to identify gaps rather than studying every service equally.
A useful readiness assessment
You are closer to the target profile when you can explain an operational choice in terms of an observed symptom and a stated requirement. You should be comfortable tracing a problem through logs, metrics, network paths, permissions, deployment state, and recovery options. You should also be able to use both the console and CLI sufficiently to understand how an operational task is performed.
Make a two-column inventory before studying. In the first column, list tasks you have performed directly: deployments, alarm investigation, access changes, backup or restore work, network troubleshooting, container operations, or infrastructure-as-code changes. In the second, list tasks you know only from reading. The second column becomes your lab and documentation priority.
Do not interpret familiarity with an AWS service as operational competence. Knowing that a service exists is different from selecting a metric, identifying a failed dependency, applying a least-privilege change, validating a deployment, or explaining how a recovery procedure affects availability. Use those actions as your readiness tests.
Which SOA-C03 domains deserve study time?
For SOA-C03, the official exam guide lists five scored domains: Monitoring, Logging, Analysis, Remediation, and Performance Optimization at 22% of scored content; Reliability and Business Continuity at 22%; Deployment, Provisioning, and Automation at 22%; Security and Compliance at 16%; and Networking and Content Delivery at 18%. Start with the three 22% domains, then close weaknesses in the remaining domains.
Monitoring, Logging, Analysis, Remediation, and Performance Optimization — 22%
This domain should be studied as an incident workflow: collect useful telemetry, interpret it, isolate the failing component, choose a proportionate remediation, and verify the result. Include the CloudWatch agent because AWS added content requiring candidates to configure and manage it to collect metrics and logs from EC2 instances, Amazon ECS clusters, or Amazon EKS clusters.
Build practice around signal selection. Ask what information would confirm a hypothesis, where that information is collected, and which action changes the condition without creating a second outage. Review alarm behavior, logs, metrics, troubleshooting approaches, and performance implications together rather than as separate product flashcards.
Reliability and Business Continuity — 22%
Study how operational procedures support availability and recovery. Map requirements such as high availability, performance, and capacity to the relevant workload behavior, then examine failure handling, backup and restore procedures, disaster recovery, and continuity decisions. The objective is not to design a complete distributed architecture; it is to implement and operate against an established requirement.
For each lab scenario, write the failure assumption first. A recovery procedure for a lost instance is not the same as one for unavailable data, a broken dependency, or a regional problem. Record what is recovered, from which source, in what order, and how you would verify service health afterward. This habit improves both troubleshooting and scenario interpretation.
Deployment, Provisioning, and Automation — 22%
This domain requires more than knowing how to launch a resource manually. Include repeatable provisioning, deployment operations, configuration changes, rollback thinking, and automation through supported AWS tools. AWS added the ability to create and manage stacks of resources using CloudFormation and the AWS CDK, so those additions should be explicit in a current study plan.
Practice comparing a manual change with an infrastructure-as-code change. Identify the declared resource, the dependency order, the update effect, the failure response, and the evidence that the intended state was reached. Include AWS Management Console and AWS CLI work so that you can recognize the operational implications of each interface.
Security and Compliance — 16%
Study security as an operational control rather than a policy-writing exercise. Review identity and access decisions, protection of workloads and data, logging and monitoring for security-relevant activity, and enforcement of stated compliance requirements. AWS added enforcing requirements such as Region and service selections, which makes constraint-aware configuration important.
When reviewing a security scenario, separate authentication, authorization, network exposure, data protection, and audit evidence. Then ask which control is being requested and which change is least disruptive while meeting it. Avoid selecting an answer merely because it sounds strongest; the best operational action must fit the stated access, workload, and compliance conditions.
Networking and Content Delivery — 18%
Prepare networking from the packet and service path outward: DNS resolution, TCP and IP behavior, routing, firewalls, connectivity, and content delivery. AWS moved VPNs from Task 4.2 to Task 5.1 in the SOA-C03 comparison, so candidates using older material should relocate that topic in their notes rather than assume the old task placement still represents the current blueprint.
AWS also added configuring and analyzing CloudWatch network monitoring services. Include network telemetry in troubleshooting practice. Draw the path from client to endpoint and identify where name resolution, routing, security rules, load balancing, or service health could interrupt it.
How is SOA-C03 different from SOA-C02?
Use SOA-C03 material if you are preparing for the current AWS CloudOps Engineer – Associate exam. AWS says SOA-C02 was in use until September 29, 2025, while SOA-C03 began September 30, 2025. The change is not only a renamed certification: the domain structure, task coverage, and emphasis changed, so an older SOA-C02 plan needs deliberate updating.
Blueprint changes to record
For SOA-C02, AWS published Monitoring, Logging, and Remediation at 20% of scored content; Reliability and Business Continuity at 16%; Deployment, Provisioning, and Automation at 18%; Security and Compliance at 16%; Networking and Content Delivery at 18%; and Cost and Performance Optimization at 12%. Those figures describe SOA-C02 and should not be presented as the SOA-C03 blueprint.
SOA-C03 combines analysis and performance optimization into Content Domain 1 and assigns that domain 22% of scored content. Reliability and Business Continuity is 22% of scored content, and Deployment, Provisioning, and Automation is 22% of scored content. Security and Compliance remains 16% of scored content, while Networking and Content Delivery remains 18% of scored content.
AWS removed configuring S3 static website hosting from Task 5.2 for SOA-C03. AWS also moved tasks 6.1 and 6.2 from SOA-C02 to Task 1.3 in SOA-C03. These changes make it risky to follow an old checklist without reading the current task statements.
When older material is still useful
Older material can help with foundational AWS operations, but validate every topic against the current SOA-C03 guide. Keep explanations that build transferable understanding of monitoring, reliability, deployment, security, networking, and troubleshooting. Mark material that focuses on removed content, old task numbering, or the former cost domain as historical or supplemental rather than current exam coverage.
What are the SOA-C03 delivery details?
AWS lists SOA-C03 as 130 minutes with 65 multiple-choice or multiple-response questions. The AWS exam guide explains that 50 questions affect the score and 15 are unscored. Results are reported as a scaled score from 100–1,000, and the minimum passing score is 720. Use these official details for planning, while checking the current booking page for any scheduling instructions that may change.
Available delivery and languages
AWS states that SOA-C03 can be taken at a Pearson VUE testing center or through online proctoring. AWS lists English, Japanese, Korean, and Simplified Chinese as available exam languages. Select the delivery method and language that match your practical circumstances, and review the current official scheduling requirements before selecting an appointment.
Do not infer that the catalogue label C90.02 carries these details automatically. Confirm that your booking is for SOA-C03. If the booking identifies SOA-C02 or gives a different exam title, stop and resolve the mismatch before proceeding.
How to use the scoring information
The presence of 15 unscored questions means you cannot know which questions affect the result. Treat every item as relevant, answer each one carefully, and avoid trying to identify unscored content. Section-level feedback should also be interpreted cautiously, as AWS warns that it is not a precise measure of mastery in every individual topic.
A scaled score is not a direct percentage of correct answers. Use the minimum passing score of 720 as the official threshold, but use practice performance diagnostically: identify the domain, task, and reasoning error behind each missed question instead of translating a practice result into a guaranteed exam outcome.
What should a four-stage study roadmap look like?
A practical roadmap moves from exam identification to foundations, then hands-on domain work, and finally timed review. Begin by confirming SOA-C03 or SOA-C02, map your experience to the current task statements, and reserve the final stage for mixed scenarios and error correction. This sequence prevents premature memorization and exposes operational gaps early.
Stage one: confirm scope and baseline
Download or read the official SOA-C03 exam guide and comparison page. Create a checklist using the five SOA-C03 domains and their named percentages. Add the target candidate requirements and the out-of-scope boundaries. Then take a baseline assessment from a legitimate practice source or your own scenario questions, recording why each answer was chosen.
Do not use leaked questions, exam dumps, or memorized answer keys as a readiness measure. They do not establish that you can operate AWS workloads and may represent an obsolete exam version. Your baseline should reveal reasoning gaps, not simply how well you recognize repeated wording.
Stage two: rebuild the operational foundation
Review the AWS Well-Architected Framework, core compute, storage, database, container, monitoring, networking, identity, and security concepts named in the official target profile. Refresh DNS, TCP, IP, firewalls, operating-system administration, scripting, Git, CI/CD concepts, and infrastructure as code as needed.
Use a service map rather than an alphabetical service list. For each workload, identify where it runs, how it is reached, how it is observed, how it is secured, how it is changed, and how it is recovered. This gives each service a job in an operational system and makes unfamiliar scenarios easier to parse.
Stage three: perform focused labs
Create small, disposable exercises that follow the domain sequence. Deploy a workload, add monitoring and logs, introduce a controlled configuration problem, investigate it, apply a remediation, and verify recovery. Repeat with an access-control issue, a network-path issue, an infrastructure-as-code change, and a backup or restore procedure.
Include the newer SOA-C03 additions directly: configure the CloudWatch agent for supported compute or container environments, create and manage a CloudFormation or AWS CDK stack, enforce a Region or service-selection constraint, and configure or analyze CloudWatch network monitoring services. Keep a short lab record containing the symptom, evidence, action, and verification result.
Stage four: integrate and decide whether to schedule
In the final review stage, work across domains instead of studying one service at a time. For every missed scenario, classify the problem as knowledge, interpretation, procedure, or careless reading. Schedule only when you can explain the correct operational action and reject plausible alternatives, not merely when your recall feels comfortable.
Use timed practice to manage attention, but do not treat it as a replica of the live exam unless the provider explicitly documents that relationship. Review the official exam details again before booking, especially the identifier, delivery option, language, and any current appointment requirements.
How should you study a difficult scenario?
Read every scenario as a set of constraints: workload, symptom, required outcome, existing controls, and limits on acceptable change. Identify what the question actually asks before evaluating services. Then eliminate answers that solve a different problem, violate the stated requirement, add unnecessary complexity, or change production without a validation or rollback path.
A repeatable reasoning method
First, name the failure or objective in one sentence. Second, identify the evidence that would distinguish the likely causes. Third, select the least disruptive action that satisfies the requirement. Fourth, check security, availability, operational effort, and reversibility. Finally, verify whether the answer describes implementation, diagnosis, recovery, or design; these are not interchangeable tasks.
For multiple-response items, count the required selections carefully and judge each option independently. A technically valid action may still be wrong if it does not address the stated condition. Do not select every answer that could be useful in a broad production environment; select the options supported by the scenario.
Where candidates commonly lose time
Candidates often spend too long comparing services before identifying the underlying requirement. Another common mistake is choosing a familiar service without checking its scope, permissions, network placement, data path, or recovery behavior. A third is treating monitoring as a dashboard exercise rather than a cycle of detection, analysis, remediation, and validation.
Create an error log with four fields: the domain, the decisive requirement, the mistaken assumption, and the rule you will use next time. Revisit the log at the end of each study session. This is more efficient than rereading every chapter after one ambiguous question.
What should you do in the final week?
The final week should consolidate decisions, not introduce an unrelated collection of services. Recheck the official blueprint, review your error log, complete a small number of mixed practice sessions, and rehearse the operational reasoning method. Keep the last review focused on weak domains and newly added SOA-C03 topics rather than chasing obscure details.
A practical final checklist
Confirm whether your appointment is for SOA-C03, not an outdated SOA-C02 entry or an unverified C90.02 label. Confirm the selected language and whether you will use a Pearson VUE testing center or online proctoring. Review the official exam duration, question format, scored and unscored question information, scaled-score range, and passing threshold.
Rehearse the five current domain names with their labels and percentages: Monitoring, Logging, Analysis, Remediation, and Performance Optimization at 22% of scored content; Reliability and Business Continuity at 22%; Deployment, Provisioning, and Automation at 22%; Security and Compliance at 16%; and Networking and Content Delivery at 18%. Keep the labels attached to the percentages in your notes.
Prepare a short explanation for each major weak area. If you cannot explain what evidence you would gather, what change you would make, and how you would verify the outcome, return to a lab or official documentation before scheduling.
What should you do next?
Your next action is to resolve the identifier, download the current official exam guide, and build a domain-based gap list. If the provider confirms that C90.02 maps to SOA-C03, use the current five-domain blueprint and updated task coverage. If it maps to SOA-C02, verify whether that exam is actually available for booking before using its older blueprint.
Three decisions before booking
Decide which official exam you are taking. Decide whether your current experience supports the target candidate profile or whether you need more hands-on practice. Decide which delivery method and language you can use comfortably. These decisions prevent the most expensive preparation mistake: studying a plausible exam that is not the one on your appointment.
After those decisions, schedule only through the official AWS certification process or the delivery provider identified by AWS. Keep the official exam guide and comparison page bookmarked so that you can check for revisions rather than relying on a static third-party catalogue entry.
Conclusion
There is no exact official AWS identifier matching C90.02 in the supplied research. The evidence points to the SOA-C02 and SOA-C03 family, with SOA-C03 now identified by AWS as AWS Certified CloudOps Engineer – Associate. Confirm the mapping first, then prepare around the current operational domains, updated task coverage, hands-on troubleshooting, and disciplined scenario reasoning. The immediate next step is not buying more questions; it is verifying the exam identifier and measuring your skills against the official SOA-C03 guide.