Amazon Aurora DSQL reaches general availability

2025-05-27 · Publisher : AWS · Domain : Cloud & DevOps

AWS declares Aurora DSQL generally available, a serverless distributed SQL database compatible with PostgreSQL, designed for 99.99% availability in a single Region and 99.999% across multiple Regions with strong consistency.

### Background On 27 May 2025, AWS announces general availability of Amazon Aurora DSQL, a distributed, serverless SQL database compatible with PostgreSQL. The service had been presented in preview at re:Invent on 3 December 2024, in three US regions. The final version adds several regions and enriches the features. ### Key facts - Active-active architecture designed for 99.99% availability in a single Region and 99.999% across multiple Regions, with strong consistency and durability of reads and writes on any Regional endpoint. - AWS claims reads and writes four times faster than other popular distributed SQL databases (vendor claim). - Disaggregated architecture into independent components: query processor, adjudicator, journal and crossbar, which scale separately based on workload. - Regions at general availability: single- and multi-Region clusters in N. Virginia, Ohio and Oregon; single-Region clusters in Osaka, Tokyo, Ireland, London and Paris. - New since the preview: improved console for multi-Region peering, PostgreSQL views, unique secondary indexes, integration with AWS Backup, PrivateLink, CloudFormation and CloudTrail. ### Why it matters Distributed databases with strong consistency and full SQL remain rare and complex to operate. Aurora DSQL targets applications that must stay available despite the loss of a Region, without managing infrastructure. For architects, it is an alternative to solutions such as Spanner or CockroachDB, with lock-in to AWS and PostgreSQL compatibility limits to check. ### Takeaway AWS brings a serverless distributed SQL database to market: multi-Region high availability becomes a managed service.

Tags : AWS, Données, Serverless, Hyperscaler