Free SAA-C03 practice: RDS vs DynamoDB decision guide

Relational or NoSQL, decided by the data — access pattern, scaling model, schema and transactions. Six comparison questions, SAA-C03, 2026 edition.

Try 6 free RDS-vs-DynamoDB questions in the exam's own style — every answer explained, no sign-up.

Straight to the 6 free questions ↓

RDS versus DynamoDB is the SAA-C03’s central database decision, and it is decided by the data, never by a vague claim that one “scales better”. Read the access pattern and the data shape, and the answer follows.

The dimensions that decide

Access pattern — varied, ad-hoc queries with joins → RDS. Known, key-based access → DynamoDB.
Scaling — vertical and read-replica scaling with a practical ceiling → RDS. Horizontal, virtually unlimited throughput, serverless → DynamoDB.
Schema — fixed, normalized, enforced → RDS. Flexible, per-item attributes that evolve → DynamoDB.
Transactions — multi-table ACID across related tables → RDS. Key-based, single-table-oriented access → DynamoDB.
Operations — managed instances you size → RDS. Serverless, pay-per-request → DynamoDB (on-demand).

Line the requirement up against these five and the choice is rarely ambiguous.

The trap: “NoSQL scales better”

The exam’s favourite distractor offers DynamoDB “because it scales better” on a workload that is fundamentally relational — joins, ad-hoc reporting, cross-table transactions. Scaling is not the deciding factor; the data model is. DynamoDB scales beautifully for key-based access, and fights you when forced into ad-hoc relational querying.

The middle option

When a workload genuinely needs relational features but also serverless scaling, the answer is not DynamoDB — it is Aurora Serverless, which keeps SQL and joins while scaling capacity automatically. Keep it in mind when a stem pairs “relational” with “scale to zero / unpredictable load”.

How to use the six questions below

Answer each before opening the explanation, and when you miss, name the dimension that decided it — “ad-hoc joins”, “horizontal scale”, “flexible schema”, “multi-table transaction”, “serverless”. Every explanation states that dimension and dismantles the distractors, including the “scales better” trap. For the domain overview, see Design High-Performing Architectures.

A decision guide comparing Amazon RDS and Amazon DynamoDB across dimensions. RDS is relational with a fixed normalized schema, SQL and ad-hoc queries with joins, multi-table ACID transactions, and vertical or read-replica scaling with a practical ceiling. DynamoDB is NoSQL with a flexible per-item schema, fast key-based access, horizontal scaling to virtually unlimited throughput, and a serverless pay-per-request operational model. The data's access pattern decides which fits.

6 free SAA-C03 practice questions

Answers and explanations — no email wall

SAA-C03 Question 1 of 6

An application's data will be queried in many different, unpredictable ways — ad-hoc reports joining several entities, filtering on arbitrary columns. Which database suits this access pattern?

Answer: A — Amazon RDS (relational).

Ad-hoc queries that join entities and filter on arbitrary columns are the relational model's strength, so RDS fits — SQL handles flexible query shapes natively. Why not the others: DynamoDB is optimized for known key-based access patterns; unpredictable ad-hoc querying fights its design, and piling on global secondary indexes to fake it is costly and still limited. S3 with Athena suits large-scale analytics, not a transactional application's varied operational queries. Flexible ad-hoc querying with joins points to RDS.

SAA-C03 Question 2 of 6

A new application must scale seamlessly to handle millions of requests per second with consistent low latency, growing horizontally without the team managing database instances or capacity. Which database fits?

Answer: A — Amazon DynamoDB.

DynamoDB scales horizontally to virtually unlimited throughput with consistent single-digit-millisecond latency and no instances to manage — the serverless scale described. Why not the others: scaling an RDS instance up (vertical) hits a ceiling and still requires managing the instance. Read replicas scale reads but not writes and add operational surface. A self-managed database on EC2 is the most operational overhead of all. Seamless horizontal scale with no instance management is DynamoDB.

SAA-C03 Question 3 of 6

A product catalog stores items whose attributes vary widely — different products have different fields, and new attributes are added often. Which database handles this flexible, evolving schema most naturally?

Answer: A — Amazon DynamoDB.

DynamoDB items can each carry different attributes, so a catalog with varying and evolving fields fits its flexible, schema-less model without migrations. Why not the others: RDS enforces a schema, so varying attributes force either sparse columns or constant ALTER TABLE migrations — friction the requirement highlights. A rigid normalized schema is the opposite of flexible. Redshift is an analytics warehouse, not an operational catalog store. Flexible, evolving per-item attributes suit DynamoDB.

SAA-C03 Question 4 of 6

An order-management system must update several related tables together in a single all-or-nothing transaction with strong consistency across them. Which database is the natural fit?

Answer: A — Amazon RDS (relational, with multi-table ACID transactions).

Coordinating updates across several related tables in one all-or-nothing transaction is the relational model's core strength, so RDS fits. Why not the others: DynamoDB supports transactions but is oriented to single-table, key-based access, and 'scales better' does not address the multi-table transactional requirement. S3 is object storage with no relational transactions. ElastiCache is an in-memory cache, not a system of record. Multi-table ACID transactions point to RDS.

SAA-C03 Question 5 of 6

A team wants a database with a fully serverless operational model — no servers to size or patch, and cost that scales to zero-ish with a pay-per-request option during unpredictable, spiky usage. Which fits without changing to a non-relational model only if necessary?

Answer: A — Amazon DynamoDB with on-demand capacity.

DynamoDB with on-demand capacity is fully serverless and bills per request, scaling with spiky, unpredictable usage and requiring no instance management — matching the operational and cost requirement (when a key-value model fits). Why not the others: a fixed large RDS instance pays for provisioned capacity around the clock. An always-on EC2 database is maximal operational overhead. RDS Multi-AZ improves availability and, if anything, adds cost — it is not a pay-per-request model. Serverless, pay-per-request scaling is DynamoDB on-demand. (If the workload truly needs relational features, Aurora Serverless is the relational serverless option.)

SAA-C03 Question 6 of 6

Which summary best captures how to choose between RDS and DynamoDB?

Answer: A — RDS for relational data, joins, ad-hoc queries and multi-table transactions; DynamoDB for key-based access at massive scale with a flexible schema and serverless operations.

The decision follows the data: relational shape, joins, flexible querying and cross-table transactions point to RDS, while known key-based access at huge scale with a flexible schema and serverless operations points to DynamoDB. Why not the others: 'scales better' and 'more familiar' are not the deciding factors — the access pattern and data model are. And they are not interchangeable, so price alone is the wrong lens. Match the data and access pattern to the database.

That is exactly how every question in the course works — answer, explanation, why-not. The real set continues in the practice player: 10 free questions, no sign-up.

Continue with the 10 free questions →

Those 6 questions were the start.

The exam does not test whether you recognise a term — it tests whether you can rule out three plausible answers under time pressure. That is what the explanations above are for, and there are 780 more questions built exactly like them.

Collecting questions yourself

Scattered across forums, of unknown age
Answer keys without reasoning
No idea which domain you are weak in

Practising with a system

780 questions in 12 full tests, SAA-C03 (2026)
Every option explained — including the wrong ones
Readiness per exam domain, and drills for your weakest

Start free — 10 questions

Straight into the player. No account, no email.

Frequently asked questions

What is the single most important factor?

The access pattern. If queries are varied and ad-hoc with joins and multi-table transactions, that is relational (RDS). If access is by key at massive scale with a flexible schema, that is DynamoDB. 'Scales better' is never the deciding argument on its own — the data model and how you read it decide.

What if I need relational features but also serverless scaling?

Consider Aurora Serverless — the relational, SQL-compatible option that scales capacity automatically. DynamoDB is the answer when a key-value/NoSQL model genuinely fits; Aurora Serverless keeps you relational while removing instance management.

Are these real exam questions?

No. They are original scenarios written for this page in the exam's style and difficulty. Reproducing real exam items violates the AWS certification agreement and can cost a candidate their certification — ours come with the reasoning attached instead.

What is the one-line takeaway?

Relational shape, joins, ad-hoc queries and multi-table transactions → RDS; key-based access at scale, flexible schema and serverless operations → DynamoDB; relational-but-serverless → Aurora Serverless.

Updated for SAA-C03 (July 2026). The sample questions above are our own work in the style of the exam — not real exam items. The exam itself is set and marked by the certification body.