# Aurora vs RDS

AWS comparisons · Updated 17 November 2026

**Short answer:** Choose Aurora for MySQL- or PostgreSQL-compatible databases that need fast failover, up to 15 low-lag read replicas, storage that grows on its own, serverless capacity or cross-Region Global Database. Choose RDS for MySQL or PostgreSQL when a single instance with a Multi-AZ standby meets your needs, the database is small or steady, or you need an engine feature or version Aurora does not offer; it is often cheaper at that scale.

## Aurora vs RDS compared

| Factor | Amazon Aurora | Amazon RDS (MySQL, PostgreSQL) |
|---|---|---|
| Engines | MySQL- and PostgreSQL-compatible | MySQL, PostgreSQL, MariaDB, Oracle, SQL Server, Db2 |
| Storage | Shared cluster volume, six copies across three AZs, grows automatically to 256 TiB | EBS-based gp2, gp3, io1, io2 per instance; you size it (storage autoscaling optional) |
| Read replicas | Up to 15, sharing the same storage | Separate instances with asynchronous replication |
| High availability | Replicas are failover targets | Multi-AZ standby (not readable) or Multi-AZ DB cluster with two readable standbys |
| Storage billing | Standard: per GB plus per I/O; or I/O-Optimized: no I/O charge, higher instance and storage rates | Per GB provisioned, plus provisioned IOPS and throughput where applicable |
| Serverless | Aurora Serverless (ACUs, can scale to zero) | No |
| Cross-Region | Global Database | Cross-Region read replicas |

## When Aurora is worth it

- You need several read replicas with low lag, because Aurora replicas read the same storage rather than replaying a log onto their own disks.
- Failover time matters and you want readers that are ready to promote.
- Storage growth is unpredictable. There is no volume to resize and no storage-full outage from forgetting to.
- Variable or dev/test workloads that fit Aurora Serverless, including scaling to zero.
- Multi-Region reads or disaster recovery with Global Database.

Pick the storage configuration deliberately: AWS recommends I/O-Optimized when I/O makes up 25% or more of your Aurora spend, and Standard below that. You can switch to I/O-Optimized once every 30 days and back to Standard at any time.

## When RDS is the better fit

- One writer and a standby is enough, and you have no heavy read scaling need.
- Small or steady databases where predictable gp3 storage costs less than Aurora's storage and I/O model.
- You need MariaDB, Oracle, SQL Server or Db2, or an extension or version not yet in Aurora.
- You want readable standbys without Aurora: RDS Multi-AZ DB clusters give a writer and two readers across three AZs for MySQL and PostgreSQL.

## Migrating between them

From RDS for MySQL or PostgreSQL you can create an Aurora read replica of the RDS instance, let it catch up, then promote it, or restore an RDS snapshot into a new Aurora cluster. Plan a short write freeze at cut-over and update the endpoint your application uses (or put RDS Proxy in front to make the switch simpler).

Going back from Aurora to RDS uses logical tools such as native dump and restore or AWS DMS. Test parameter groups and extensions: Aurora versions track community versions but are not identical.

## Common wrong reasons

- **"Aurora is always faster."** AWS quotes up to 6x the throughput of stock MySQL or PostgreSQL on similar hardware; your query plan and indexes still decide most real performance.
- **"Aurora is always more expensive."** For read-heavy clusters with several replicas it can cost less, because replicas share storage.
- **"RDS Multi-AZ gives me a read replica."** A Multi-AZ DB instance standby cannot serve reads; use a read replica or a Multi-AZ DB cluster.

## Questions

### Is Aurora more expensive than RDS?

Per instance-hour Aurora usually costs more, and Aurora Standard adds I/O charges. With several replicas or I/O-heavy workloads on I/O-Optimized, total cost can be similar or lower. Compare with your own usage on both pricing pages.

### Is Aurora PostgreSQL the same as RDS PostgreSQL?

Both run PostgreSQL-compatible engines, but Aurora uses a different distributed storage layer and supports a slightly different set of versions and extensions. Test before migrating.

### Can I migrate from RDS to Aurora without downtime?

With an Aurora read replica of the RDS instance you can keep downtime to a short cut-over while you promote the replica and repoint the application.

## When this happens away from your desk

Database incidents are usually connections exhausted, storage filling or a failover in progress; knowing which within a minute changes the response. CloudConsole is being built to show RDS and Aurora alarms and events on a phone with production guardrails.

[About the CloudConsole app](https://cloudconsole.dev/aws-console-app) · [Download on the App Store](https://apps.apple.com/app/cloudconsole-for-aws/id6835419710)

## Sources

- [What is Amazon Aurora?](https://docs.aws.amazon.com/AmazonRDS/latest/AuroraUserGuide/CHAP_AuroraOverview.html)
- [Amazon Aurora storage](https://docs.aws.amazon.com/AmazonRDS/latest/AuroraUserGuide/Aurora.Overview.StorageReliability.html)
- [Amazon RDS DB instance storage](https://docs.aws.amazon.com/AmazonRDS/latest/UserGuide/CHAP_Storage.html)

## Related

- [RDS Multi-AZ vs read replica](https://cloudconsole.dev/vs/rds-multi-az-vs-read-replica)
- [Aurora Serverless v2 vs provisioned](https://cloudconsole.dev/vs/aurora-serverless-v2-vs-provisioned)
- [DynamoDB vs RDS](https://cloudconsole.dev/vs/dynamodb-vs-rds)
- [Savings Plans vs Reserved Instances](https://cloudconsole.dev/vs/savings-plans-vs-reserved-instances)
- [RDS instance in storage-full state](https://cloudconsole.dev/errors/rds-storage-full)
- [How to fail over an Aurora cluster](https://cloudconsole.dev/how-to/failover-rds-aurora-cluster)

---

Source: https://cloudconsole.dev/vs/aurora-vs-rds

CloudConsole is an independent iPhone and Android app for Amazon Web Services by Cloudconcur Ltd. It is not affiliated with Amazon Web Services.
