C90.02 Exam Guide: Identify the AWS Version Before You Prepare
C90.02 does not appear as an exact official AWS certification identifier in the supplied AWS documentation. The closest matches are SOA-C02, the previous AWS Certified SysOps Administrator - Associate exam, and SOA-C03, its updated AWS Certified CloudOps Engineer - Associate version. Both validate practical AWS operations skills rather than service memorization alone. This guide helps you decide which blueprint applies to your goal, check whether your study material is obsolete, and build preparation around deployment, monitoring, security, networking, reliability, and operational troubleshooting.
What does C90.02 refer to?
Treat C90.02 as an unverified catalogue label until you confirm it with the organization that published the listing. AWS identifies the previous exam as SOA-C02 and the updated exam as SOA-C03; the supplied official research found no exact AWS certification identifier matching “SOA C90.02.”
paragraphs2
Choose between the previous and current AWS blueprint
If your registration, course, or practice package specifically names SOA-C02, check its status before studying. AWS states that SOA-C02 was the previous AWS Certified SysOps Administrator - Associate exam, while SOA-C03 is the updated AWS Certified CloudOps Engineer - Associate exam. AWS lists September 29, 2025 as the final day to take SOA-C02 and September 30, 2025 as the first day to take SOA-C03.
Do not assume that a question bank labelled C90.02 is current merely because its title includes AWS operations terminology. Compare its domain names and task coverage with the official AWS exam guide. A package built around six SOA-C02 domains is materially different from one built around the five SOA-C03 domains, especially in monitoring, reliability, automation, and performance optimization.
The practical decision is straightforward: if you need a currently available AWS certification, investigate SOA-C03 through AWS Certification before booking or purchasing preparation material. If an employer or internal catalogue explicitly requires the older code, ask that organization whether it means SOA-C02 or another vendor’s exam. Do not pay for a preparation product until the code, version, and testing organization agree.
What the current exam validates
SOA-C03 validates the ability to deploy, manage, and operate workloads on AWS. Its official task description includes operating workloads according to the AWS Well-Architected Framework, using the AWS Management Console and AWS CLI, applying security controls, monitoring and troubleshooting systems, handling networking concepts, implementing high availability and performance requirements, performing continuity and disaster recovery procedures, and identifying and remediating incidents.
This is an operations exam. Your preparation should therefore connect each service to an operational decision: what signal indicates failure, which control limits exposure, which deployment action is reversible, which network path is permitted, and how a recovery procedure restores service. Reading isolated service summaries is less useful than tracing a workload through deployment, observation, failure, remediation, and verification.
AWS also lists several boundaries for the target role. The candidate is not expected to design distributed architectures, design CI/CD pipelines, design hybrid and multi-VPC networking, develop software, define security or governance requirements, develop ransomware defense strategies, assess and plan resource capacity, analyze costs and total cost of ownership, or manage business continuity and disaster recovery strategies. Use those boundaries to prevent study time from drifting into architect- or developer-level design work.
Who should take this exam?
The intended candidate has approximately one year of experience deploying, managing, troubleshooting, networking, and securing AWS workloads, along with at least one year in a related operations role such as system administrator. Candidates without that background can still study the material, but they should expect to build operational fluency rather than rely on short-term recall.
The official general knowledge list includes monitoring, logging, and troubleshooting techniques; DNS, TCP, IP, and firewall concepts; high availability, performance, and capacity concepts; at least one scripting language; a major operating system; cloud computing; containerization and orchestration; CI/CD and Git. AWS also expects familiarity with storage and container solutions, monitoring tools, CloudFormation, infrastructure as code, networking and security services, hybrid and multi-VPC operations, databases, and compute services such as Amazon EC2, AWS Lambda, and Amazon ECS.
Use this list as a readiness check. If you can explain an incident but cannot identify the relevant metric or log, begin with observability. If you know the console but not the CLI or infrastructure as code, add repeatable operations. If networking terms are unfamiliar, study the request path before memorizing AWS service features. The largest gaps should determine your sequence.
Which SOA-C02 domains matter for a legacy requirement?
For a genuine SOA-C02 requirement, study the six official scored domains and keep the domain name attached to every weighting. Monitoring, Logging, and Remediation accounts for 20% of scored content; Reliability and Business Continuity accounts for 16%; Deployment, Provisioning, and Automation accounts for 18%; Security and Compliance accounts for 16%; Networking and Content Delivery accounts for 18%; and Cost and Performance Optimization accounts for 12%.
Monitoring, Logging, and Remediation should anchor the plan because it is the largest SOA-C02 domain. Cover the flow from signal collection to diagnosis, action, and confirmation. Reliability and Business Continuity requires a different perspective: failure modes, recovery objectives, resilient components, backup handling, and controlled restoration. Deployment, Provisioning, and Automation focuses on repeatability and safe change.
Security and Compliance should be studied as operational controls rather than abstract policy. Networking and Content Delivery requires you to reason about DNS, routing, connectivity, filtering, and delivery behavior. Cost and Performance Optimization is smaller by the official weighting, but it should not be ignored; understand how operational choices affect efficiency and service behavior.
Do not compare the percentages as if they were interchangeable scores. Each percentage describes scored content within its named SOA-C02 domain. AWS reports results as a scaled score rather than as a simple percentage, so domain weighting is a prioritization tool, not a promise that mastering one domain produces a predictable result.
How does SOA-C03 change the study emphasis?
SOA-C03 reorganizes the blueprint into five content domains. Monitoring, Logging, Analysis, Remediation, and Performance Optimization is 22% of scored content; Reliability and Business Continuity is 22% of scored content; Deployment, Provisioning, and Automation is 22% of scored content; Security and Compliance is 16% of scored content; and Networking and Content Delivery is 18% of scored content.
The change is not just a new title. AWS added CloudWatch agent configuration for collecting metrics and logs from EC2 instances, Amazon ECS clusters, or Amazon EKS clusters. AWS also added CloudFormation and AWS CDK stack management, compliance enforcement such as Region and service selections, and CloudWatch network monitoring services. VPNs moved from the security area to networking, while content from the former SOA-C02 cost and performance tasks moved into Task 1.3.
AWS removed configuring S3 static website hosting from Task 5.2. That does not mean S3 is irrelevant to all AWS operations, but it does mean older material may spend time on a task no longer listed in the SOA-C03 comparison. Before using SOA-C02 notes, mark every addition, deletion, and recategorization in the official comparison document.
For a current SOA-C03 objective, distribute time across the three 22% domains rather than carrying the SOA-C02 assumption that cost and performance optimization is a separate 12% domain. Preserve foundational cost reasoning where it supports performance optimization, but follow the current task structure.
Build a lab-first preparation strategy
Use a short theory-to-action loop for every topic: read the task statement, configure or inspect a small AWS scenario, create a failure or policy change, observe the result, and document the remediation. This method tests whether you can operate a workload, not merely recognize a service name in a multiple-choice question.
Start with one simple workload and make it observable. Record the expected health signals, relevant logs, dependencies, and recovery action. Then change one condition at a time: remove a permission, alter a security rule, create an unhealthy target, change a route, or introduce a deployment configuration error. The goal is to practice diagnosis without relying on memorized answer patterns.
For each lab, write five lines: symptom, evidence, likely cause, corrective action, and verification. Add the AWS service or control involved. This produces a revision record that exposes weak reasoning. For example, if your corrective action is “check CloudWatch,” it is incomplete; specify which metric or log, what threshold or event matters, and what result would confirm the diagnosis.
Use the AWS Management Console and AWS CLI where the task requires operational familiarity. Use infrastructure as code for repeatable resources when the topic involves provisioning or automation. The AWS guide specifically names AWS CloudFormation and, for SOA-C03, AWS CDK stack management. You do not need an elaborate production platform; a small reproducible environment is more valuable than a large lab that you cannot troubleshoot.
A practical six-stage study roadmap
A useful roadmap moves from baseline knowledge to operational integration. Do not assign identical study time to every service. Begin with the blueprint version, then prioritize domains by their official weighting and by your own incident, networking, automation, and security gaps.
Stage one: identify the exam. Save the exact code shown by AWS or the authorized testing organization. If the listing says C90.02, request clarification and compare it with SOA-C02 and SOA-C03. Confirm whether the objective is a current certification or a legacy requirement before selecting books, courses, or question banks.
Stage two: map the objectives. Create a table with one row per domain and task. In the next column, write the AWS services and concepts named by the official guide. In another, record whether you can explain, configure, troubleshoot, and verify the topic. A topic is not ready because you have read it; it is ready when you can support an operational decision with evidence.
Stage three: establish the foundations. Review IAM principles, compute, storage, databases, VPC networking, DNS, logging, monitoring, containers, the Well-Architected Framework, CLI use, and infrastructure as code. Keep the review focused on operations: permissions, health, dependencies, scaling behavior, recovery, and controlled change.
Stage four: complete domain labs. Work through monitoring and remediation first for SOA-C02, or the corresponding 22% Monitoring, Logging, Analysis, Remediation, and Performance Optimization domain for SOA-C03. Then study reliability, deployment, security, and networking. If you are following SOA-C03, explicitly include CloudWatch agent work, CloudFormation or CDK stacks, compliance enforcement, and CloudWatch network monitoring.
Stage five: integrate scenarios. Combine domains in incidents such as an application that is reachable but returning errors, a deployment that creates unhealthy targets, a workload that cannot access a dependency, or a recovery process that restores data but not application connectivity. Explain the order of checks and why each action is safe.
Stage six: perform a readiness review. Revisit only missed objectives and concepts you cannot explain without notes. Use practice questions as diagnostic prompts, not as a substitute for the official blueprint or hands-on work. For every missed item, identify whether the failure came from service knowledge, requirement parsing, elimination, or operational reasoning.
How to study scenario questions without exam dumps
Practice by translating each scenario into requirements before looking at answer choices. Extract the workload, symptom, constraint, desired outcome, and prohibited side effect. Then eliminate options that solve a different problem, weaken the stated control, introduce unnecessary operational effort, or fail to verify the result.
A strong review record includes the correct decision, the evidence supporting it, and the reason the alternatives fail. If a question concerns an access problem, separate identity permissions from resource policies, network reachability, encryption, and application configuration. If it concerns an outage, separate detection, diagnosis, mitigation, recovery, and prevention.
Do not use leaked questions, exam dumps, or memorized answer keys as a preparation strategy. They can leave you unable to handle changed wording and may conflict with the current blueprint. Practice material is most useful when it sends you back to an official task statement or a lab that proves the underlying behavior.
For multiple-response practice, ask whether every selected option satisfies the full requirement. One technically valid action may still be incomplete if the scenario asks for both remediation and verification, or for the least disruptive operational change. This habit is more durable than remembering a phrase from an unofficial question set.
Common preparation mistakes to avoid
The most expensive mistake is studying the wrong version. A C90.02 label can conceal a legacy SOA-C02 reference or an incorrect catalogue entry. Resolve the identifier first, because domain coverage, task placement, and current availability differ between SOA-C02 and SOA-C03.
Another mistake is treating the exam as a service catalogue. Knowing what a service does is not the same as selecting it under a constraint. Add the missing operational context: availability requirement, recovery need, access boundary, network path, deployment method, monitoring signal, or performance symptom.
Avoid spending the entire plan on the most familiar service. Candidates often over-study EC2, S3, or IAM while neglecting DNS, logs, container operations, automation, recovery, or network troubleshooting. The official target profile is broader than one compute-and-storage workload.
Do not mistake a practice score for proof of readiness. Review the reasoning behind every answer, including correct guesses. Section-level feedback also needs caution; AWS explicitly warns that it should not be overinterpreted. Use it to locate study areas, not to infer a precise capability measurement.
Finally, do not turn out-of-scope architecture work into a distraction. The official guide separates operating an existing workload from designing distributed architectures, CI/CD pipelines, or hybrid networking. Learn enough design context to operate safely, then return to the actions and evidence expected from the CloudOps role.
What are the current SOA-C03 delivery details?
The supplied AWS source lists SOA-C03 as a 130-minute exam with 65 multiple-choice or multiple-response questions. AWS states that it can be taken at a Pearson VUE testing center or through online proctoring, and lists English, Japanese, Korean, and Simplified Chinese as available exam languages.
AWS reports SOA-C03 results as a scaled score from 100–1,000 and lists 720 as the minimum passing score. The exam includes 50 questions that affect the score and 15 unscored questions that do not affect the score. Because the score is scaled, do not treat the number of correct answers as a directly published passing percentage.
These details apply to SOA-C03, not automatically to a label called C90.02 or to the previous SOA-C02 exam. Confirm the version and booking information on the official AWS Certification page before scheduling. Delivery availability, registration conditions, and other booking information should be checked at the point of scheduling rather than copied from an old preparation page.
How should you decide when to schedule?
Schedule only after you can map the correct version’s tasks to evidence from labs, notes, or work experience. You should be able to explain an incident from symptom to verification, use the relevant AWS operational tools, and identify why an attractive alternative violates the scenario’s security, reliability, networking, or automation requirement.
Use a simple readiness gate. First, every blueprint domain has a completed review. Second, the highest-priority domains have hands-on scenarios rather than reading-only notes. Third, you can work through unfamiliar wording without searching for an answer key. Fourth, your remaining errors have a known cause and a plan for correction.
If your experience is below the official target profile, allow additional time for operating-system, networking, scripting, container, and cloud fundamentals. If you already administer AWS workloads, spend less time on definitions and more time on cross-service incidents, CLI usage, infrastructure as code, monitoring evidence, and recovery procedures.
Before booking, perform one final identity check: exam code, exam name, version, official delivery page, language, and testing route. This protects you from preparing for a catalogue label that does not correspond to the certification you intend to earn.
Your next actions
The next action is not downloading more questions; it is resolving the code and building a version-specific objective map. Once the target is confirmed, use the official AWS exam guide as the authority, the comparison page to identify version changes, and practical labs to turn each task into an operational skill.
Write down whether your target is SOA-C02, SOA-C03, or an unresolved C90.02 listing. For SOA-C03, review the five current domains and its added monitoring, automation, compliance, and network-monitoring content. For a verified SOA-C02 legacy requirement, use the six-domain weighting and confirm that the organization still accepts that credential.
Then choose a first lab: collect workload metrics and logs, investigate a controlled failure, apply the least disruptive remediation, and verify recovery. Continue by connecting that evidence to deployment, security, networking, and reliability decisions. This sequence gives your study plan a concrete starting point and keeps preparation aligned with the work the AWS CloudOps certification is designed to validate.
Conclusion
C90.02 should be treated as a label requiring verification, not as an official AWS exam code confirmed by the supplied research. The decisive preparation step is to identify whether the requirement means SOA-C02 or the current SOA-C03. After that, follow the matching blueprint, prioritize its named domains, practise operational scenarios, and schedule only after confirming the official exam and delivery details. That approach reduces the risk of studying an obsolete outline and builds the troubleshooting judgment the certification is intended to assess.