Apache Flink 1.19 - Deprecations, New Features, and Improvements

Apache Flink 1.19 - Deprecations, New Features, and Improvements

Mar 18, 2024

Apache Flink® 1.19 is here! On behalf of the Flink community, David Anderson highlights key release updates with FLIPs for Legacy deprecations, Flink SQL, Observability, Flink Configuration, and Flink Connectors.

Legacy config methods and options
► FLIP-381: Deprecates configuration methods that use complex Java objects
► FLIP-391: Deprecates RuntimeContext#getExecutionConfig
► FLIP-398: Improves serialization configuration and usage In Flink
► FLIP-405: Migrates string configuration key to ConfigOption
► FLIP-416: Deprecates and removes RestoreMode#LEGACY

Legacy SQL stack
► FLIP-392: Deprecates the Legacy Group Window Aggregation
► FLIP-395: Deprecates Global Aggregator Manager

Flink SQL
► FLINK-24024: Adds support for SESSION windows to capture and analyze user behavior patterns based on their activity sessions.
► FLINK-20281: Allows for the use of input streams like Apache Kafka with Debezium and Upsert Kafka when you want to apply a window aggregation.
► FLIP-400: Introduces the new UDF type, AsyncScalarFunction. This allows the issuing of concurrent function calls that improve performance.
► FLIP-373: Enables using SQL hints to configure state TTL for regular streaming joins and group by aggregations.

► FLIP-384: Enables users to create Trace Reporters in order to export spans to an external backend.
► FLIP-385: Adds MetricReporter and TraceReporter implementations using OpenTelemetry.
► FLIP-386: Enables state backends to attach custom attributes during recovery to recovery spans

Flink Configuration
► FLIP-366: Introduces support for standard YAML. It eliminates syntax confusion for users and enables users to leverage third-party tools for maintaining the Flink configuration.

Flink Connectors
► FLIP-372: Introduces mixin interfaces to enhance the extendibility of the Sink API.

Many more FLIPs are a part of this release. See the blog post for more details.

► See release notes for AF 1.19: https://nightlies.apache.org/flink/flink-docs-release-1.19/release-notes/flink-1.19/
► Read the blog post: https://cnfl.io/3TIoZKt
► Download AF 1.19: https://flink.apache.org/downloads/
► Get started with AF 1.19: https://nightlies.apache.org/flink/flink-docs-master/docs/dev/table/sql/gettingstarted/


00:00 - Intro

00:43 - FLIP deprecations

00:59 - FLINK-24024: Support Session Window TVFs

01:50 - FLINK-20281: Window aggregation supports changelog input stream

02:07 - FLIP-400: AsyncScalarFunction for asynchronous scalar function support

02:22 - FLIP-373: Configure Different State TTLs using SQL Hints

03:01 - FLIP-384: Introduce TraceReporter - create checkpointing and recovery traces

03:09 - FLIP-366: Support standard YAML for Flink configuration

03:33 - FLIP-372: Enhance Sink API

03:56 - It’s a wrap!

Confluent is pioneering a fundamentally new category of data infrastructure focused on data in motion. Confluent’s cloud-native offering is the foundational platform for data in motion – designed to be the intelligent connective tissue enabling real-time data, from multiple sources, to constantly stream across the organization. With Confluent, organizations can meet the new business imperative of delivering rich, digital front-end customer experiences and transitioning to sophisticated, real-time, software-driven backend operations. To learn more, please visit www.confluent.io.

#apacheflink #flink #confluent