In today’s big data field, stream processing is becoming increasingly popular, with Apache Flink being a rising star that has caught everyone’s attention. However, the 24/7 operational challenges it brings cannot be ignored. In the current context of cost reduction and efficiency improvement, the effective utilization of resources has become a key focus.

This presentation will provide a detailed explanation of the Flink K8S Operator, a sub-project derived from the Apache Flink community. It will briefly introduce the origins and development history of this project, as well as the automatic tuning function introduced in the latest version. From the current community implementation of automatic tuning (FLIP-271) function to a detailed explanation of its working principle and best practices, the presentation will also touch upon the community’s ongoing implementation of non-stop updates (FLIP-291). Finally, it will discuss some future plans for automatic tuning work in the Flink community and at StreamPark.

Speakers:

Chen Zhengyu Apache Flink/Streampark Contributor, a volunteer in the Apache Flink Chinese community, has been engaged in data development in the game industry for a long time, and is currently in charge of Flink’s public multi-cloud data solutions, data synchronization, and the design and construction of big data operation pipeline platforms from 0 to 1 in a game company.