PostgreSQL 11

PostgreSQL 11 expanded declarative partitioning with hash partitioning, default partitions, partition-wise joins and aggregation, and the ability to move rows between partitions during updates. It added SQL procedures invoked with CALL, including transaction control when the call context permits it. The release introduced just-in-time expression compilation for some queries, added INCLUDE columns to B-tree indexes for covering index-only scans, and broadened parallel-query capabilities. It also improved operational behavior in areas such as parallel index creation, partition maintenance, and monitoring.

PostgreSQL 11 has reached end of life, so it no longer receives community security fixes, bug fixes, or compatibility corrections. Remaining on it leaves known and newly discovered defects unpatched and makes it harder to obtain support from extension vendors, managed-service providers, and operating-system platforms as they retire older builds. The longer an installation remains on this major version, the more likely its upgrade will span several release generations, increasing the chance of extension incompatibilities, deprecated configuration assumptions, query-plan changes, and application behavior that has not been tested against current PostgreSQL releases.

Plan a move to a currently supported PostgreSQL major version and test the target version with a representative copy of production data and workload. Inventory extensions, procedural languages, client drivers, backup tooling, replication tooling, authentication settings, and operating-system packages before selecting the target. Run schema and application compatibility tests, compare important query plans and performance, and validate backup, restore, failover, and rollback procedures. Choose an upgrade method appropriate to the allowed downtime and database size: pg_upgrade can perform an in-place major-version upgrade, while logical dump and restore or logical replication can support a more controlled migration with different cutover characteristics. Rebuild or refresh optimizer statistics after cutover, verify extension versions and index health, and retain a tested rollback path until production behavior is confirmed.