Amazon has recently announced the general availability of Amazon Aurora DSQL, a PostgreSQL-compatible, serverless, and distributed database. The new managed service is designed to support active-active high availability and multi-region strong consistency.
Announced in preview during the latest re:Invent keynote, Aurora DSQL targets use cases that demand global scalability, resilience, and consistent performance. Channy Yun, lead blogger at AWS, writes:
Unlike most traditional databases, Aurora DSQL is disaggregated into multiple independent components such as a query processor, adjudicator, journal, and crossbar. These components have high cohesion, communicate through well-specified APIs, and scale independently based on your workloads.
In a single region, Aurora DSQL provides 99.99% availability, while multi-region deployments are designed for 99.999% availability, with the system built to recover automatically from failures. Two regional endpoints act as peers, supporting concurrent operations while maintaining strong consistency, while the third region functions as a log-only witness to enhance availability. Yun adds:
Aurora DSQL is an ideal choice to support applications using microservices and event-driven architectures (…) It’s also ideal for multi-tenant software as a service (SaaS) applications and data-driven services like payment processing, gaming platforms, and social media applications.
When deploying multi-region clusters, all read and write operations are consistent and durable across all regional endpoints, with each regional endpoint presenting a single logical database. This enables applications to read from and write to any location.
Source: AWS blog
Luc van Donkersgoed, Principal Engineer at PostNL and AWS Hero, writes:
This is the biggest product launch in years, rivaling Lambda and DynamoDB in its potential.
In the article “Just make it scale: An Aurora DSQL story,” Niko Matsakis and Marc Bowes, both senior principal engineers at AWS, provide technical insights on Rust and how they used it to build DSQL. They write:
The goal with Aurora DSQL’s design is to break up the database into bite-sized chunks with clear interfaces and explicit contracts. Each component follows the Unix mantra—do one thing, and do it well—but working together they are able to offer all the features users expect from a database (transactions, durability, queries, isolation, consistency, recovery, concurrency, performance, logging, and so on).
Source: All Things Distributed
Marc Brooker, VP and distinguished engineer at AWS, has previously published an in-depth set of posts about the DSQL design and implementation and the thinking behind the new service.
While the reaction in the community has been generally positive, some users are confused by the new normalized billing unit called Distributed Processing Unit (DPU) and how to compare it with existing alternative offerings, as well as compatibility concerns.
AWS claims that “Aurora DSQL and PostgreSQL return identical results for all SQL queries,” and the new database supports core relational features such as ACID transactions, secondary indexes, joins, inserts, and updates. However, the documentation provides a detailed list of SQL feature compatibility, with a section about unsupported PostgreSQL features. User sh1boleth comments:
I think for orgs that already use full-fledged MySQL/Postgres this won’t offer many benefits other than ease of use and infinite scaling, however for orgs that use DynamoDB and feel restricted with its modelling and lack of relations (and don’t wanna handle the headache that is Single Table) this is a nice middle ground between a full relational DB vs NoSQL.
With the GA launch, the service has added support in additional regions in Europe and Asia Pacific and introduced support for PostgreSQL views, CloudFormation, KMS customer-managed keys, and AWS Backup. A Model Context Protocol (MCP) server for AI application workloads, which converts human-readable questions and commands into structured PostgreSQL-compatible SQL, has been announced as part of the GA.
Similar to other managed AWS services, there is a free tier that covers 100000 DPUs and 1 GB-month of storage each month.
