Standard support for Amazon Aurora PostgreSQL-Compatible Edition and Amazon RDS for PostgreSQL version 13 will end on February 28, 2026. PostgreSQL 13 will be deprecated by the community in November 2025, ceasing to receive bug fixes or security patches. AWS recommends upgrading to newer versions, such as 16 or 17, which offer significant performance enhancements and improved security. PostgreSQL 17 can achieve up to twice the write throughput and consumes 20 times less memory during vacuum operations. Version 16 introduces pg_stat_io for detailed I/O statistics, while version 14 includes a vacuum emergency mode. Aurora-specific enhancements in version 14.9 and later can lead to faster query latency and reduced costs.
Version 14 introduces new roles for access control, and version 15 revokes certain permissions. Major upgrades in logical replication include automatic slot synchronization in version 17 and support for parallel apply in version 16. Transitioning between major versions requires careful examination of catalog changes, as some views and configuration parameters will evolve. Extensions must be verified, as most do not auto-upgrade.
An in-place major version upgrade can be performed via the AWS Console or CLI, with downtime varying based on database size. AWS recommends snapshot-based testing beforehand. The CLI command can check valid upgrade targets, leading from version 13 to 14, 15, 16, or 17. Preparation involves validating instance classes and dropping replication slots.
Amazon RDS Blue/Green deployments allow for near-zero downtime by synchronizing production with a staging environment, enabling application testing before traffic switching. This feature is supported from Aurora PostgreSQL version 13.12 onward. Logical replication through pglogical offers flexibility for minimal downtime, while AWS DMS supports homogeneous migration with Change Data Capture.
Extended Support is available for a fee, providing up to three years of security patches. Best practices include replicating production environments in staging, conducting load tests, and validating queries against new catalogs. Recent minor releases, including Aurora PostgreSQL 17.6 and 16.10, showcase ongoing improvements. Engaging AWS Support is advisable for complex setups to ensure seamless transitions before the deadline.