{{ examPhase === 'setup' ? 'Practice setup' : examPhase === 'result' ? 'Round score' : 'Round progress' }}

{{ summaryValue }}

{{ summaryLine }}

Unofficial AWS Certified Solutions Architect – Professional · SAP-C02

Guide verified 12 September 2026. Independently authored; not affiliated with or endorsed by AWS. Practise organizational, resilience, security, migration and cost trade-offs.

24 curated English scenarios, six per domain, with a 12-item multi-constraint subset. Rounds sample uniformly without replacement. These author-assigned scenarios are not calibrated exam difficulty or full syllabus coverage.

The actual exam has 75 questions in 180 minutes and includes multiple-response items. This practice is untimed and single-answer; it does not assess hands-on architecture skills. Official weights are D1 26%, D2 29%, D3 25%, D4 20%; sampled rounds do not enforce these weights.

Official guide and domains · Eligibility and exam format

Optional. Blank uses sap-c02-practice. A custom seed changes question and option order; up to 128 UTF-16 units are used without splitting Unicode characters.

{{ seedNote }}

{{ examSetupValidation }}

{{ answeredCount }} of {{ questionCount }} answered · {{ Math.round(progressPercent) }}%

{{ progressStatusLine }}

Question {{ examCursor + 1 }} of {{ questionCount }}

{{ currentQuestion.stem }}

{{ currentQuestion.accessible_clue }}

Answer recorded. Corrections appear when you view results.

Review this practice round

{{ resultLead }}

SAP-C02 · guide verified 12 September 2026 · original bank 2026.09.13.1. AWS and its service names are trademarks of Amazon.com, Inc. or its affiliates.

This score covers only this authored round. It does not predict passing or map to an AWS scaled score. This 24-item bank is untimed, uniformly sampled and not difficulty-calibrated; it omits multiple-response and hands-on assessment and does not cover the full syllabus. Use the official guide and practical architecture work alongside these explanations.

{{ readinessReviewTitle }}

{{ readinessReviewLine }}

{{ card.label }}
{{ card.value }}
{{ card.note }}

Missed answer review

  1. {{ item.prompt }}

    Your answer: {{ item.selected_answer_label }}

    Correct answer: {{ item.correct_answer_label }}

    {{ item.explanation }}

    {{ item.source.label.split(' — ')[0] }} · {{ item.source.version }}

Domain performance

Correct and missed answers by official domain for this round. Domains absent from the chart were not sampled. Use the missed-item sources to review that domain, then practise a focused set and check the full official guide.

Loading domain-performance chart…

{{ examChartStatus }}

Choose your next domain

Official weights describe the real exam, not this round. A domain with no sampled answers is untested, not a zero score. Use the full AWS objectives alongside this small practice bank.

Answer ledger

Question ID Attempt identity Prompt Topic Your answer Correct answer Outcome Explanation Source Copy
{{ row.question_id }} {{ attemptIdentity }} {{ row.prompt }} {{ row.category_label }} {{ row.selected_answer_id }} · {{ row.selected_answer_label }} {{ row.correct_answer_id }} · {{ row.correct_answer_label }} {{ row.outcome_label }} {{ row.explanation }} {{ sourceUrl(row.question_id) }}{{ row.source ? row.source.version : '' }}
AWS Solutions Architect Associate Exam Prep Practise AWS Solutions Architect Associate questions with 32 original SAA-C03 scenarios and explanations for focused domain review.
Cloud Egress Cost Calculator Estimate cloud egress spend from measured traffic and a marginal rate card with verified reductions, peak-month stress testing, and budget variance.
Cloud NAT Gateway Cost Calculator Estimate monthly cloud NAT cost from gateway time and routed traffic while comparing offload savings, route charges, and budget impact.
Object Storage Cost Calculator Estimate monthly object storage cost from capacity and request volume with retrieval, egress, lifecycle and editable provider rate assumptions.

Two network connections can still fail together when they terminate at the same Direct Connect location. Protecting against a failed device and protecting against a failed location require different designs. Architecture on Amazon Web Services (AWS) depends on identifying that distinction, then choosing a solution that also meets the workload’s security, cost and operational requirements.

AWS Certified Solutions Architect – Professional SAP-C02 examines architectural decisions across complex organizations, new applications, existing workloads, and migration and modernization projects. AWS describes the intended candidate as having two or more years of experience designing and implementing AWS solutions. That is recommended experience, not a claim that a practice score can establish equivalent skill.

Requirements often pull in different directions. Business units may need freedom to administer their accounts while central teams retain control of audit protections. A recovery environment may need to serve traffic promptly without carrying the expense of full production capacity all year. For an existing application, increasing capacity before measuring the bottleneck can leave the original problem unsolved.

Recovery planning separates two requirements that are easy to confuse.

Recovery time objective (RTO)
The maximum acceptable delay between service interruption and service restoration.
Recovery point objective (RPO)
The maximum acceptable data-loss window, expressed as time. Recent changes that have not reached a recovery copy can fall inside that window.

A warm standby keeps a functioning application running at reduced capacity in another AWS Region. Recovery can then concentrate on scaling and redirecting traffic. A pilot-light design keeps core resources available but needs additional actions before it can serve requests. Backup and restore requires recovering data and rebuilding the necessary environment. These approaches have different running costs and recovery work, and their names do not guarantee an RTO or RPO. Recovery exercises must establish what the whole application can actually achieve.

Migration has a similar distinction between moving a workload and changing its design. Rehosting moves supported servers with limited application changes. Modernization can replace parts of a monolith or change the database and application architecture. A successful data copy alone does not prove that application behavior, dependencies or rollback arrangements will work after cutover. The appropriate answer depends on the deadline, permitted changes and failure consequences stated in the scenario.

Scenario practice is useful when the explanation reveals why an apparently reasonable alternative breaks a requirement. Recognizing an answer after repeated exposure is weaker evidence than being able to explain the decision and test it in a controlled environment. Educational practice does not certify professional competence or guarantee an examination pass.

Exam version also matters. As of 10 October 2026, AWS has announced 16 November 2026 as the last day to take SAP-C02. Registration for SAP-C03 opens on 27 October 2026, with general availability beginning on 17 November 2026. Preparation based on SAP-C02’s four-domain outline should not be assumed to cover the replacement examination.

How to Use This Tool:

Use these independently authored English scenarios to review architectural reasoning. This practice is unofficial, is not affiliated with or endorsed by AWS, and supports SAP-C02 rather than SAP-C03.

  1. Choose Domain / practice focus. Use All four domains · uniform sample for mixed practice, one of D1 to D4 for focused study, or Multi-constraint review · all domains for the selected cross-domain subset.
  2. Choose the Question count. The initial mixed round contains eight questions, with up to 24 available. A domain contains six questions and the multi-constraint subset contains 12; their menus offer smaller round sizes. Changing focus can reset a count that the new focus does not support.
    If Start exam practice is disabled, choose an available focus, feedback mode and question count from the menus.
  3. Choose Feedback timing. The default, Explain after each answer, shows the correction, explanation and reference after a choice. Review explanations at the end records each choice but hides correctness until results. Both modes are untimed.
  4. Enter an optional Replay seed under Advanced. A blank entry uses sap-c02-practice. Surrounding spaces are removed. Long entries are shortened at a whole character to fit the allowance of 128 basic text characters; many emoji use two places. Reproducing a round requires the same focus, count, feedback mode, seed and unchanged practice content.
  5. Select Start exam practice, read the constraints, and choose one of the four answers. Selecting an answer immediately records and locks it. Use Next question to continue and View results after the final answer.
    End round abandons the current attempt and returns to setup. It does not produce a final score for an unfinished round.
  6. Review Missed answer review and the explanations in the Answer ledger. Follow a missed question’s reference to check its reasoning, then use Retry missed for those questions or Practise domain to start a six-question domain round.
    If Domain performance cannot draw its chart, the domain counts remain available. Use Retry chart to request the chart again.

Interpreting Results:

During an attempt, Round progress counts answers recorded, regardless of correctness. After completion, Round score is the number correct out of all questions in that round. Perfect round means every selected answer matched the authored key. Neither label establishes readiness, predicts passing, or verifies the ability to configure, secure or recover an AWS workload.

The Domain performance chart compares counts of correct and missed answers, not official domain weights. A domain with more questions has more opportunities for both correct answers and misses. Not sampled in this round means no evidence was collected for that domain, not that its score is zero. Use the individual explanations and the full official objectives to decide what needs further study.

Replay same seed repeats the current round’s questions and option order. Try a new seed uses a different seed, but familiar questions can appear again. A full 24-question mixed round already uses the entire bank, so another full-bank round cannot add unseen questions. These are repeatable practice samples, not adaptive difficulty or independently calibrated tests.

Retry missed starts a separate round restricted to the current round’s incorrect answers, potentially only one question. Its score describes that smaller retry, not a revised score for the earlier round. Compare like-for-like attempts and distinguish improved reasoning from recognition of previously revealed answers.

Technical Details:

A single-best-answer architecture question fixes a set of requirements and asks which offered design satisfies them. An alternative can name a suitable AWS service yet fail because it omits a permission, return route, consistency assumption or migration prerequisite. Practice scoring records agreement with the authored answer key; it does not evaluate a deployed architecture.

Points, accuracy and completion

Each correct answer earns one point and each incorrect answer earns zero. Let N be all questions in the current round, A the number answered, and C the number correct. The following equations distinguish points, accuracy, progress and misses.

Score points = C Accuracy (%) = 100×CN Progress (%) = 100×AN Missed answers = A−C

Possible points equal N, so the score ranges from zero to N. There is no partial credit, negative marking or domain multiplier. Unanswered questions remain in the accuracy denominator but are not counted as incorrect or missed. Full review requires every question to be answered.

The main score uses exact integer counts. Raw accuracy is retained without rounding. The visible progress percentage is rounded to the nearest whole percent, with half-percent ties rounded up, so two recorded answers in a six-question round show 33% progress regardless of correctness. Progress reaches 100% when all answers are recorded, including when some or all are wrong.

Condition Meaning Review outcome
A < N At least one question is unanswered. The round remains in progress; no completed review is available.
A = N and C < N All questions are answered and at least one is wrong. Start with the missed answers; missed-item retry is offered.
A = N and C = N Every question is correct. Perfect round; there are no missed questions to retry.

These rules describe completion and review, not pass/fail bands. A domain’s counts use only its sampled questions. An absent domain contributes no attempted questions and no evidence about performance.

Official weighting and authored coverage

The bank contains 24 original English scenarios, six assigned to each SAP-C02 domain. The table distinguishes official examination weights from the subjects represented by those six-question practice pools.

SAP-C02 domain Official weight Subjects in the practice pool
D1 · Design Solutions for Organizational Complexity 26% Organization permission limits, transit routing, hybrid name resolution, federated account access, multi-account template rollout, and cost allocation tags.
D2 · Design for New Solutions 29% Recovery environments, DynamoDB consistency, reliable event publication, private CloudFront origins, cross-account encryption-key access, and Direct Connect location failures.
D3 · Continuous Improvement for Existing Solutions 25% Private Amazon Simple Storage Service (S3) transfer paths, compute commitments, storage performance, Aurora read connections, staged releases, and protected backups.
D4 · Accelerate Workload Migration and Modernization 20% Database change replication, server rehosting, file transfers, schema conversion, incremental replacement, and dependency-based migration waves.

Selection is uniform without replacement within the chosen pool: a question cannot appear twice in one round, but no official domain quota is enforced. All 24 questions give six per domain, or 25% each, rather than the official proportions. Short mixed rounds can omit domains altogether.

The multi-constraint pool contains 12 of those same questions, three from each domain, not 12 additional questions. Its grouping and the domain assignments are author-selected study aids. They do not establish official difficulty, exhaustive task coverage or validated domain proficiency.

The authored material identifies a guide verification snapshot of 12 September 2026 and question-bank version 2026.09.13.1. These identify the material being practised; the verification date is not an examination launch date or a guarantee that every explanation remains current as AWS services change.

Conditions that change an architectural answer

An AWS Organizations service control policy (SCP) limits permissions; it does not grant them. In an organization with all features enabled, an applicable explicit deny can restrict ordinary principals in descendant member accounts even when an AWS Identity and Access Management (IAM) policy allows the action. Management-account principals and service-linked roles are outside that SCP restriction. An account-administration answer must therefore identify whose actions are constrained, retain the necessary permission grants, and account for exceptions.

DynamoDB multi-Region eventual consistency (MREC) uses asynchronous replication and last-writer-wins conflict resolution for concurrent updates to the same item. It does not promise that every field change survives competing writes. The scenario explicitly selects MREC; its conclusion must not be generalized to the separate multi-Region strong consistency (MRSC) mode. Data ownership or conflict-aware application logic matters when losing an update is unacceptable.

A transactional outbox addresses a different consistency problem: a database change and a separately published event can succeed or fail independently. Committing the business record and its pending event in one local transaction removes that gap. A relay publishes committed events, but retries can deliver duplicates. Consumers must be idempotent, meaning that processing the same event again does not repeat its business effect. Publishing before the database commit can instead announce a change that never commits.

For an Aurora cluster with healthy replicas, the reader endpoint distributes new read connections across those replicas. It does not balance each query within a single persistent connection. The reporting scenario therefore depends on replica-lag tolerance and the ability to use multiple connections; adding replicas without changing where the reporting service connects does not resolve the stated writer bottleneck.

Migration answers also depend on what must move. AWS Transform MGN, called AWS Application Migration Service in older material, supplies server replication and test/cutover launches for supported rehosting workloads. AWS Database Migration Service (DMS) handles supported database migrations. Its full-load-plus-change-data-capture (CDC) approach copies existing data and applies subsequent changes, but requires the relevant source logs, engine support, replication monitoring and a controlled cutover. Neither service name removes the need to validate application behavior.

Why this is not an official exam simulation

SAP-C02 has 75 questions within 180 minutes: 65 affect the score and 10 are unidentified unscored questions. Official multiple-choice items have one correct response among four choices. Multiple-response items require all correct responses to be selected from five or more choices. These practice rounds use only four-choice, single-answer questions and have no timer, multiple-response assessment or hands-on architecture assessment.

AWS reports SAP-C02 results on a scaled range of 100 to 1,000, with a minimum passing score of 750. That scale accounts for differences between examination forms; it is not a percentage-correct scale. A 75% practice result must not be converted to 750, and no raw-score conversion or pass prediction is supplied. Review the official guide for the version you will actually take and combine question review with practical architecture work.

Privacy Notes:

Scoring happens in your browser rather than through a separate answer-grading service. However, the current address encodes the attempt, seed, responses and position without encryption. Someone who receives that address can restore a compatible attempt and its answers. The address can remain in browser history and can reach the site when loaded again.

Keep credentials and private information out of Replay seed. Shared addresses and exported answer ledgers disclose responses. Ending a round clears its state from the current address, not from previously copied links, browser history or saved documents.

If a link cannot be restored because its saved attempt is invalid or incompatible with the current content or saved-attempt format, Saved attempt unavailable directs you to start a replacement round. A warning that the attempt could not be saved in the URL means it remains open but will be lost on reload; end the round and start again to recover URL saving.

Chart and document support can load external resources, and opened AWS references use the network. Browser-based scoring is therefore not a promise of fully offline operation or of an attempt never leaving the device.

Worked Examples:

One missed transaction in a mixed round

With question-bank version 2026.09.13.1, select the all-domain focus, four questions, end-of-round feedback and the seed sap-literal. This round contains schema-conversion, transactional-outbox, private-origin and organization-boundary questions. Suppose only the outbox answer is wrong. The completed score is 3/4, equivalent to 75% accuracy. D1 and D4 each have one correct answer; D2 has one correct and one missed answer; D3 was not sampled. Review the outbox explanation before selecting Retry missed. That retry contains one question even though the ordinary setup menus start at two. Correcting it earns 1/1 in the retry, not a revised 4/4 for the earlier round.

Shared services without application-to-application routing

Select all six D1 questions with immediate explanations to include the transit-gateway scenario. Its two application virtual private clouds (VPCs) must reach shared services without routing to each other, their address ranges do not overlap, and security groups already allow the required service traffic. The accepted choice associates application attachments with a table routing only to shared services and uses a separate shared-services table with return routes to both applications. Selecting it earns one point and reveals the routing explanation. Propagating every route into one common table would remove the required routing isolation, while omitting return routes would break replies. In a practical exercise, check the VPC subnet routes as well as the transit-gateway tables rather than treating the accepted choice as a complete deployment configuration.

References: