Published on Oct 18, 2022
In addition to improved sorting and compression, MERGE and other workload management tools are included in the most popular open-source database.
With the release of the latest version of PostgreSQL, performance improvements have been made over PostgreSQL 14 as well as new functions and capabilities for managing workloads in both local and distributed environments.
PostgreSQL 15 was released on October 13 by the PostgreSQL Global Development Group, which oversees the development of the database. It can be downloaded from the PostgreSQL website.
As public cloud service providers such as Oracle, Google, and Microsoft add PostgreSQL support to their products as a result of its increasing popularity, the new capabilities become increasingly important.
The Microsoft Cosmos DB recently added PostgreSQL support, and Google Cloud Platform (GCP) launched AlloyDB in May, a PostgreSQL-compatible, fully managed database-as-a-service (DBaaS).
In its new version, PostgreSQL has improved in-memory and on-disk sorting algorithms, the development group announced. Benchmark tests show a 25%-to-500% speed increase, depending on the type of data.
Developers sort query output tables to ensure that they do not appear in an unspecified order. As a result, the database management system will return an output table based on the scan and order inside the disk, which is not useful for analysis.
In the new version of PostgreSQL, row_number(), rank(), dense_rank(), and count() can be used as window functions to provide performance benefits. Developers can use window functions to perform calculations across a set of rows associated with the current query.
Additionally, queries using SELECT DISTINCT — an SQL statement that removes duplicate rows from the result table — can now be executed in parallel in the new version of PostgreSQL.
Despite the fact that many queries cannot be executed in parallel, those that can can be executed more than twice as fast using this method, according to the development team.
Based on the previous version of PostgreSQL, which supported async remote queries, the new version of PostgreSQL now supports asynchronous commits through the implementation of a new tool, the PostgreSQL foreign data wrapper.
Contrary to synchronous code architecture, where tasks are dependent upon each other (performed sequentially), asynchronous architecture allows tasks to run independently and does not cause the code to stop working or to freeze when one task fails.
The new archiving and backup facilities are also expected to improve performance.
According to the group, PostgreSQL 15 now supports LZ4 and Zstandard (zstd) data compression algorithms, which provide space and performance benefits for certain workloads.
As part of the new version of the database management system, custom modules can also be used for archiving, eliminating the need to use shell commands.
