The development team behind Apache Flink, the framework for processing data streams, has released version 1.16 of the framework. Users now have access to a number of interesting new features thanks to the active participation of over 240 contributors.

Since its launch, Apache Flink has become the de facto standard for both stream and batch processing. The combination of the two capabilities makes it an excellent choice to power various kinds of analytics applications, including the most demanding ones.

The highlights of the release include overall improvements in performance, stability, and usability for batch processing. With the introduction of SQL Gateway extension and full compatibility with Hive Server2, users can submit Flink SQL jobs and Hive SQL jobs very easily, and it is also simple to connect to the original Hive ecosystem.

There are a number of significant improvements for stream processing, the most notable being Changelog State Backend, which now provides users with precise checkpoints to significantly improve fault tolerance experience while also offering a smaller end-to-end latency experience for transactional Sink jobs.

More detailed information can be found in the official announcement of Apache Flink 1.16 on the Apache Software Foundation blog.