Apache Flink is an open-source distributed processing engine with support for stateful processing and event time semantics. It offers interfaces for stream and batch processing in multiple programming languages like Java, Python, Scala, and SQL.
Amazon Managed Service for Apache Flink now supports Apache Flink version 1.19.1, released by AWS community, bringing bug fixes and enhancements. The new release introduces various features and configuration changes for Managed Service for Apache Flink users.
The latest version includes improvements in connectors, SQL API, state TTL per operator, session window table-valued functions, mini-batch optimization for regular joins, and AsyncScalarFunction. It also supports Python 3.11 and dynamic checkpoint interval adjustments for better performance.
Apache Flink 1.19 enables users to configure state Time-To-Live through SQL hints, simplifying the process without manual JSON plan manipulation. The release also enhances session window TVFs and introduces mini-batch processing for regular joins to reduce the load on the state backend.
AsyncScalarFunction in Apache Flink 1.19 supports non-blocking calls for asynchronous I/O in streaming mode. The new release also supports Python 3.11 and improves checkpoint behavior by allowing dynamic interval adjustments. Checkpoint and job initialization traces are now easier to access in logs for troubleshooting.
Managed Service for Apache Flink has undergone behavior changes in handling runtime configuration parameters when modified programmatically. With runtime 1.19, explicit exceptions are raised if configuration changes conflict with managed service settings.
The latest release of Apache Flink 1.19 brings numerous enhancements for developers, focusing on performance, flexibility, and usability. Managed Service for Apache Flink now supports the latest Flink version, offering users the option to upgrade their applications seamlessly.
For further details on Apache Flink releases and updates, users can refer to the Apache Flink blog and release notes. New users are encouraged to explore the right API and language for their needs and start using Managed Service for Apache Flink with the getting started guide.
Article Source
https://aws.amazon.com/blogs/big-data/amazon-managed-service-for-apache-flink-now-supports-apache-flink-version-1-19/