RDS vs Aurora: Differences and Which the AWS Exam Expects (2026)

Illustration of a single dark database cylinder beside a cluster of three connected database cylinders glowing orange

Amazon Aurora is a MySQL- and PostgreSQL-compatible database that AWS built for the cloud, with shared storage across three Availability Zones, faster failover, and features like Global Database and Serverless v2. Amazon RDS is the managed service for running standard database engines, including Oracle, SQL Server, MariaDB and Db2, which Aurora doesn’t support. On the exam, pick Aurora when the scenario stresses performance, many low-lag read replicas, fast failover or cross-Region reads, and pick RDS when it names an engine Aurora can’t run or asks for a simple, standard setup.

Facts verified November 2026 against AWS documentation.

RDS vs Aurora at a glance

In short: choose RDS for standard engines and simple setups, and Aurora for MySQL or PostgreSQL workloads that need more performance, faster failover or cross-Region reads.

Aurora is technically one of the engines you launch through the RDS console and API, but it works very differently underneath. The exam treats them as separate choices.

Amazon RDS (standard engines)Amazon Aurora
EnginesMySQL, PostgreSQL, MariaDB, Oracle, SQL Server, Db2MySQL-compatible and PostgreSQL-compatible only
StorageProvisioned EBS storage per instance. Optional storage autoscalingOne shared cluster volume that grows automatically, up to 256 TiB (Aurora FAQs)
Data copiesMulti-AZ adds a standby copy in another AZ6 copies of your data across 3 AZs, always
Read replicasUp to 15 (MySQL, MariaDB, PostgreSQL), asynchronous, each with its own storageUp to 15 Aurora Replicas sharing the cluster volume, typically low replica lag
FailoverMulti-AZ instance deployment: typically 60–120 secondsTypically within 30 seconds, to a replica you can prioritise
Cross-RegionCross-Region read replicasAurora Global Database: up to 10 secondary Regions, typical replication lag under 1 second
ServerlessNoAurora Serverless v2 scales capacity up and down with demand
BackupsAutomated backups and point-in-time restoreSame, plus Backtrack for Aurora MySQL (rewind without a restore)

Choose RDS when…

  • The question names Oracle, SQL Server, MariaDB or Db2. Aurora can’t run them. This single word settles the question.
  • You’re lifting and shifting a standard database and want the same engine, with AWS handling backups, patching and failover.
  • The workload is small or steady and doesn’t need Aurora’s extra availability or performance features.
  • The application needs engine-specific features or versions that only the community or commercial engine provides.

Choose Aurora when…

  • The scenario needs high read throughput with many replicas and very little replica lag.
  • Failover must be fast, measured in seconds rather than minutes.
  • Users are spread across Regions and need low-latency local reads, or the business needs cross-Region disaster recovery with an RPO of seconds and an RTO of about a minute. That’s Aurora Global Database.
  • Traffic is unpredictable or intermittent, and the question asks for capacity that scales automatically. That’s Aurora Serverless v2.
  • Storage growth must be hands-off. The Aurora volume grows on its own.
  • The question says “MySQL- or PostgreSQL-compatible” with “high performance” or “high availability”. These words point straight to Aurora.

Exam cues: which answer the question wants

Cue in the questionLikely answerExam
“Managed relational database service” with no other detailAmazon RDSCLF-C02
“AWS-built, MySQL and PostgreSQL compatible, high performance”Amazon AuroraCLF-C02
“Oracle” or “Microsoft SQL Server” with “managed”Amazon RDS for Oracle or SQL ServerCLF-C02, SAA-C03
“Read replicas with minimal replication lag”, “up to 15 replicas”Aurora ReplicasSAA-C03
“Regional outage, RPO of seconds, RTO under a minute”Aurora Global DatabaseSAA-C03
“Global users, low-latency reads in each Region”Aurora Global DatabaseSAA-C03
“Unpredictable, spiky workload”, “scale capacity automatically”Aurora Serverless v2SAA-C03
“Undo a bad write quickly without restoring from backup” (MySQL)Aurora BacktrackSAA-C03, SOA-C03
“Survive the loss of an AZ” on a standard engineRDS Multi-AZ deploymentSAA-C03, SOA-C03
“Offload reporting queries from the primary”Read replicas (RDS or Aurora; check the engine)SAA-C03, SOA-C03

Worked example 1: the reporting slowdown

A company runs a MySQL database on Amazon RDS. Analysts run heavy reports during the day and the web application slows down. The company wants to keep the reports and needs the reporting data to be as current as possible. It’s willing to change database engine if that helps. What should a solutions architect recommend?

Start with the problem: reads from the reports are competing with the application. Any read replica would move the reports off the primary. So the real question is which replica gives the freshest data. Standard RDS read replicas use asynchronous replication to separate storage, so lag can grow under load. Aurora Replicas read from the same shared cluster volume as the writer, so their lag is typically much lower. The phrase “as current as possible” plus “willing to change engine” points to migrating to Aurora MySQL and pointing the reports at an Aurora Replica through the reader endpoint. If the question had said “with the least change”, an RDS read replica would be the better answer.

Worked example 2: the Regional outage

A financial company needs its database to keep working if an entire AWS Region becomes unavailable. It can lose no more than a few seconds of data and must be back online within minutes. The database is PostgreSQL. Which solution meets these requirements?

Read the two numbers. An RPO of seconds rules out restoring from snapshots copied between Regions: snapshots are periodic, so you’d lose hours of data. An RTO of minutes rules out rebuilding anything by hand. A cross-Region read replica on RDS for PostgreSQL could work, but replication lag and a manual promotion make the targets harder to meet. Aurora PostgreSQL with Aurora Global Database replicates storage to the secondary Region with lag that’s typically under a second, and a secondary Region can be promoted in under a minute. That’s the answer the exam expects. If the question had said Oracle, Aurora would be out, and you’d look at RDS cross-Region options instead.

Common RDS vs Aurora traps

  1. “Aurora supports all RDS engines.” It doesn’t. Only MySQL and PostgreSQL compatibility.
  2. Multi-AZ isn’t read scaling. On a standard RDS Multi-AZ instance deployment, the standby doesn’t serve reads. Use read replicas for that. (Aurora Replicas do both jobs: they serve reads and act as failover targets.)
  3. Read replicas aren’t backups. A replica copies a bad delete just as faithfully as a good write. Use automated backups and point-in-time restore, or Backtrack on Aurora MySQL.
  4. “Serverless” means Aurora, not standard RDS. If the question wants a relational database that scales capacity automatically, standard RDS isn’t the answer.
  5. DynamoDB isn’t an Aurora alternative for relational needs. If the question mentions joins, SQL or an existing MySQL app, stay relational.

RDS vs Aurora FAQ

Is Aurora part of RDS?

Yes, in the sense that you create and manage Aurora through Amazon RDS. But its storage, replication and failover work differently, so AWS and the exams treat it as its own service.

Is Aurora faster than RDS MySQL?

AWS states that Aurora delivers up to five times the throughput of standard MySQL and up to three times that of standard PostgreSQL. Real results depend on the workload.

Is Aurora more expensive than RDS?

Often, per instance-hour. But Aurora can cost less overall when it replaces several RDS replicas or when Serverless v2 lets capacity shrink during quiet periods. For the exam, “most cost-effective” with a small, steady workload usually favours standard RDS.

Can I move from RDS MySQL to Aurora MySQL?

Yes. AWS supports migrating from RDS for MySQL or PostgreSQL to Aurora, for example by creating an Aurora read replica of the RDS instance and promoting it.

Which one appears more on the CLF-C02 exam?

CLF-C02 questions stay at recognition level: know that RDS is managed relational databases across several engines and Aurora is AWS’s high-performance MySQL- and PostgreSQL-compatible engine. The detailed trade-offs above are SAA-C03 material.

Test yourself

Database questions appear throughout the associate exams. Check yourself on the Solutions Architect format:

Free mock exam

Take SAA-C03 Mock Exam 1, free

A full-length, timed mock in the real exam format, with every answer explained so you know exactly where you stand.

65

questions

130

minutes

720

to pass

Free account required.

Then go further with our SAA-C03 practice exams. Studying for a specific exam? Read the CLF-C02 exam guide, the SAA-C03 exam guide or the SOA-C03 exam guide.


AWSBoy is not affiliated with Amazon Web Services. AWS and related marks are trademarks of Amazon.com, Inc. or its affiliates.

Sources (checked November 2026): Amazon Aurora FAQs · Amazon Aurora DB clusters · Using Amazon Aurora Global Database · Failing over a Multi-AZ DB instance · Amazon RDS FAQs (read replicas)

Shopping Cart
Scroll to Top