What the AWS SAA exam asks you to do

The AWS Certified Solutions Architect (Associate), or AWS SAA, is a proctored, timed test with multiple-choice and multiple-response questions. The questions use short architecture scenarios and requirements. You need to choose appropriate AWS services and patterns, interpret constraints such as security, high availability, disaster recovery, performance and cost, and identify the best option among plausible alternatives.

That means a useful AWS SAA study plan is not just a tour of service descriptions. Practise translating each requirement into a design decision: which service fits, what settings or architecture pattern matter, and why the alternatives do not meet the constraints as well. Prioritise security and least privilege, reliability, managed services where suitable, cost-performance trade-offs, multi-AZ design, and networking, routing and DNS.

This plan runs for six weeks at four two-hour sessions per week: eight hours weekly. Treat that as a workable starting schedule, not a promise that a certain number of hours guarantees readiness. If you have less time, keep the order and reduce the scope of each session. Before you begin, check the current AWS exam guide and official AWS documentation for the exam scope and any service details you need to verify.

Set up your study board

Build a board from your own notes, AWS documentation, course material and practice-question explanations. Organise it around decisions you must make in scenarios, not just a catalogue of service names. Keep a section for distinctions that are easy to confuse, such as choosing between S3 storage classes or between queueing and streaming services.

A MySummaries board on this topic could organise those notes like this:

AWS Solutions Architecture AWS SAA architecture decisionsStudy
Associate exam preparationAWS SAA architecture decisions5 sections · 3 columns
Availability and recovery
  • Multi-AZ RDS — synchronous standby for availability; standby is not a read-scaling target
  • S3 cross-Region replication — copies eligible objects asynchronously; configure versioning on source and destination buckets
Networking and access
  • Security groups are stateful, instance or ENI-level controls
  • Network ACLs are stateless subnet-level controls
  • Route 53 alias records can point to supported AWS resources
Storage selection
  • S3 — object storage for durable access to objects
  • EBS — block storage attached to EC2 in one Availability Zone
  • EFS — managed shared file storage for Linux workloads across multiple AZs
Cost and operations

Prefer managed services when they satisfy the requirements and reduce operational work. Check the required availability, throughput, latency and data-retention needs before choosing a cheaper design.

Decoupling and events
  • SQS — decouple producers and consumers; standard queues are at-least-once
  • SNS — publish a message to multiple subscribers
  • Kinesis Data Streams — ordered streaming records within a shard

A repeatable two-hour session

Use the same rhythm through the plan so each session produces evidence of learning rather than just time spent reading.

  • First 15 minutes: retrieve the previous session’s key decisions from memory. Check your notes only after attempting an answer.
  • Next 45 minutes: study one topic using your source material. Add service limits, configuration choices and trade-offs only when they help answer a scenario.
  • Next 40 minutes: answer scenario questions. For each, identify the requirement, constraints, best-fit design and strongest competing option. Explain why the competing option falls short.
  • Final 20 minutes: review errors and update the board. Turn repeated confusions into short recall prompts and schedule them for later review.

The due queue gives you a quick way to decide whether to start with recall or new material. The board names below are example study areas, not a recommended AWS exam weighting.

A MySummaries due queue for a candidate’s boards might look like this:

Due today
AWS SAA architecture decisions12 due
AWS networking and routing7 due
AWS data and storage servicesClear

The six-week schedule

Each week has four two-hour sessions. Keep one rest day between sessions where possible. If you miss a session, resume with the next planned topic and use the next review block to catch up; avoid trying to cram several missed sessions into one sitting.

Week 1: Requirements, compute and storage

Session 1 — Scenario reading and baseline (2 hours). Work through a small set of mixed practice scenarios. For every answer, underline the stated requirements and constraints, then record the service or pattern you chose and the reason. Use the results to make a short list of topics to revisit; do not treat a single score as a readiness verdict.

Session 2 — Compute (2 hours). Compare EC2, Lambda and container options such as ECS and EKS at the level required to choose between them: execution model, operational responsibility, scaling needs and workload fit. Practise spotting when a scenario calls for event-driven execution, persistent compute or container orchestration.

Session 3 — Storage (2 hours). Review S3, EBS, EFS and FSx against access pattern, attachment, shared access and durability needs. Add lifecycle, versioning and replication distinctions when they are relevant to the scenario. Use the final practice block to explain why the closest alternative is less suitable.

Session 4 — Review and mixed practice (2 hours). Revisit the baseline errors. Make a one-page decision map for compute and storage, then answer a fresh set of questions without looking at the map until you have selected an answer.

Week 2: Networking and traffic routing

Session 1 — VPC foundations (2 hours). Draw a VPC with public and private subnets across more than one Availability Zone. Label route tables, internet gateway, NAT gateway and security boundaries. Be able to trace how traffic reaches its destination rather than memorising a diagram.

Session 2 — Security and connectivity (2 hours). Compare security groups with network ACLs, then review private connectivity choices such as VPC endpoints and VPN or Direct Connect use cases. Check current AWS documentation when a configuration detail or service limitation matters.

Session 3 — DNS and load balancing (2 hours). Practise matching Route 53 routing policies and load balancer choices to the stated routing, health-check and application needs. Explain what happens when a target becomes unhealthy and which component makes the routing decision.

Session 4 — Network scenarios (2 hours). Answer scenario questions focused on private access, routing failures and cross-environment connectivity. For each error, write the missing network path or security rule in one sentence.

Week 3: Databases, integration and application patterns

Session 1 — Relational databases (2 hours). Study RDS and Aurora, including Multi-AZ availability and read-replica use cases. Distinguish failover from read scaling; do not assume a standby is serving reads.

Session 2 — NoSQL and caching (2 hours). Compare DynamoDB and ElastiCache by access pattern and role. Practise reading requirements for low-latency key-value access, relational querying, caching and data consistency before choosing a service.

Session 3 — Messaging and event patterns (2 hours). Compare SQS, SNS and Kinesis Data Streams. For each scenario, ask whether it needs a work queue, fan-out to subscribers or an ordered stream of records. Note delivery and ordering details only as they apply to the selected service.

Session 4 — Build and critique (2 hours). Sketch a simple web application using compute, storage, database and traffic routing. Add a failure path and explain how the design handles it. Then answer mixed questions on databases and integration.

Week 4: Security, resilience and disaster recovery

Session 1 — IAM and data protection (2 hours). Practise least-privilege reasoning with IAM roles, policies and resource access. Review encryption choices for data at rest and in transit, and identify where keys are managed when the scenario specifies that requirement.

Session 2 — Resilience (2 hours). Compare single-AZ and multi-AZ designs, health checks, load balancing and managed-service failover. For every design, state which component fails and what continues working.

Session 3 — Backup and disaster recovery (2 hours). Compare backup and restore, pilot light, warm standby and multi-site approaches in terms of recovery needs, complexity and cost. Select only after identifying the required recovery objectives in the scenario; do not assume a target that was not given.

Session 4 — Security and reliability practice (2 hours). Work through mixed questions and explain the most important security and failure-mode considerations before you choose an answer. Revisit any answer where you selected a service but could not describe its role.

Week 5: Performance, scaling and cost

Session 1 — Scaling and throughput (2 hours). Review Auto Scaling, load balancing, caching and service-specific scaling approaches. Distinguish a bottleneck in compute from one in storage, database or network throughput.

Session 2 — Cost decisions (2 hours). Practise comparing a managed option with a more operationally intensive alternative. Include data transfer, storage access patterns and idle capacity in your reasoning where the scenario makes them relevant. Do not optimise cost by violating availability or performance requirements.

Session 3 — Mixed timed practice (2 hours). Complete a timed practice block using the timing and question style of the real test as represented in your current official preparation materials. Afterward, review every uncertain response, not just incorrect ones. The real exam is timed, but check the current official exam guide rather than relying on an assumed question count or duration.

Session 4 — Repair a weak area (2 hours). Use your practice results to select the weakest topic. Spend one hour rebuilding the relevant decision map and one hour answering questions on it. A topic picker can direct the next session towards weak cards and lost marks.

A MySummaries topic picker might point the candidate to this area for the next focused review:

New lecture
How long have you got?5 min10 min15 min20 min
VPC routing and private accessRDS availability choicesS3 storage patterns

VPC routing and private access · Struggling — missed 5 of 12 recent recall prompts and selected an incorrect route or endpoint choice in recent practice

Generate lecture ideas

Week 6: Consolidate and check readiness

Session 1 — Mixed practice (2 hours). Answer a fresh timed set that mixes services and constraints. Practise selecting the best fit, not merely identifying a service that could work.

Session 2 — Error analysis (2 hours). Sort mistakes into knowledge gaps, misread requirements, overlooked constraints and weak comparison with the distractor. Review the relevant source material, then write a corrected explanation in your own words.

Session 3 — Architecture walkthroughs (2 hours). Take three short scenarios and use a consistent answer skeleton: requirements and constraints; proposed architecture; security and IAM; reliability and recovery; performance and scaling; cost and operations. Tailor the headings to the scenario and name the AWS services and settings that matter.

Session 4 — Light review (2 hours). Review due cards, your highest-error topics and service distinctions that you still confuse. Stop adding broad new material. Check the official exam guide and your booking details for any current requirements, then plan a normal night’s rest before the test.

After each week, compare performance by topic rather than focusing only on an overall percentage. A weak-area report can show whether your next study block should address networking, storage or another topic. Its “times” count should be read as your practice exposure, not as a claim about how often the official exam tests that subject.

A MySummaries weak-area report might rank practice results like this:

Where marks go missing
52%VPC routing and endpoints12×
67%Database availability and replicas9×
84%S3 storage and replication10×

How MySummaries helps

MySummaries can turn your AWS notes and course material into a revision board, then use that board for recall cards and written practice. For this plan, keep the board organised around service choices, constraints and trade-offs; use due reviews and weak-topic results to decide what to practise next.