This new function additionally introduces a extra environment friendly technique of merging incremental backups offline into full backups, decreasing the server load usually required for a full backup. Designed and applied by EDB worker and long-time Postgres hacker and committer Robert Haas, this performance is predicted to be standard amongst DBAs, particularly these managing on-premises Postgres deployments. Moreover, the function features a strong set of hooks that allow storage to many alternative targets, in addition to help for options like compression and tarball storage.

New in-core backup performance means sooner restoration instances, decreasing restoration time goals (RTO) dramatically—a necessary requirement for enterprises operating massive databases or high-availability programs. This makes Postgres an much more engaging possibility for organizations the place downtime has important enterprise impacts, similar to these dealing with large-scale analytics and AI workloads.

SQL/JSON is a part of the SQL:2023 commonplace, offering database builders with a strong set of features for interacting with JSON paperwork in a means that feels pure to these accustomed to SQL. Postgres has been progressively enhancing its SQL/JSON performance over the previous a number of releases, however Postgres 17 marks a major leap ahead. It introduces extremely requested options like JSON_TABLE(), together with question features similar to JSON_EXISTS(), JSON_QUERY(), and JSON_VALUE(), and a wealthy new set of constructors for constructing JSON SQL objects.