PostgreSQL 17 integrates better with JSON and requires fewer query rewrites

PostgreSQL 17 integrates better with JSON and requires fewer query rewrites

PostgreSQL version 17 is available and promises significant improvements in performance, scalability, and ease of use. The popular open-source database, known as Postgres for short, now offers improved integration with the JSON standard common in NoSQL databases.

The new release of Postgres focuses on attracting more enterprise-oriented database management. Among the key features, as mentioned, are the enhanced JSON capabilities, which will allow Postgres to compete with document-based database systems such as MongoDB. The relational model underlying Postgres remains intact, meaning more flexibility for companies that want to use both.

Maintenance of the more than 35-year-old database standard is in the hands of The PostgreSQL Global Development Group. More than 450 developers and companies contribute to this project run by volunteers. Jonathan Katz is one of them. At VentureBeat, he explains how the latest update improves logical replication, making Postgres more reliable for production environments. Memory management has also been significantly enhanced. This is thanks to a more efficient vacuum process, which frees up space after rows in the database are deleted. “PostgreSQL 17 makes vacuum more efficient and performant.” Katz said.

Walking through the index more efficiently

Another notable performance improvement in PostgreSQL 17 has to do with query execution. Queries using the ‘IN’ clause with B-tree indexes now require less manual rewriting of queries. Databases such as PostgreSQL use B-tree indexing to organize data to search the whole thing faster. This works somewhat like how an index in the back of a book allows the reader to quickly find what page a topic is on, instead of having to flip through the entire book.

Before this update, database administrators sometimes had to rewrite how they asked for data to speed up searches (a query rewrite). PostgreSQL 17 now automates these types of queries when the “IN” clause (asking: ‘is a value in a column?’) is used with B-tree indexes, without the need for rewriting. Users only need to upgrade to the new version, and the database handles it more efficiently behind the scenes.

Other improvements include native support for incremental backups. Although extensive third-party support for this feature already exists, it is now also possible by default.

Also read: EDB Postgres AI is a new platform for analytics workloads