Dear community,
this is the weekly community update thread #47. Please post any news and updates you want to share with the community to this thread.
# Updates on sharing state between subtasks
Jamie opened a first PR to add a first version of sharing state between tasks. It works by using the JobMaster as the point of synchronization [1].
# Refactor source interface
There is a lively discussion about refactoring Flink's source interface to make it future proof [2]. Join the discussion if you want to learn more.
# Task speculative execution for batch jobs
Tao Yangyu started a discussion about executing batch tasks speculatively in order to mitigate the straggler problem [3]. If you have good ideas for this problem, then please chime in.
# 2nd release candidate for Flink 1.7.0
The community just published the second release candidate for Flink 1.7.0 [4]. Please help the community by testing the latest release candidate and report any encountered problems. Thanks a lot!
# Embracing Table API in Flink ML
Weihua kicked off a discussion about how to build Flink's next Machine Learning pipelines [5]. He drafted a design document with his proposal. Join the discussion to learn more about and share your opinion.
# Support for interactive programming in Flink Table API
Jiangjie started a discussion about interactive programming with Flink's Table API [6]. The idea is to make results of previous jobs accessible to successive Flink jobs to better support a REPL like job execution.
Cheers,
Till
Locked
