You are viewing a plain text version of this content. The canonical link for it is here.

Posted to dev@flink.apache.org by "Viktor Feklin (Jira)" <ji...@apache.org> on 2021/12/28 04:15:00 UTC

[jira] [Created] (FLINK-25462) Allow override partitioning for KeyedStream

Viktor Feklin created FLINK-25462:
-------------------------------------

             Summary: Allow override partitioning for KeyedStream
                 Key: FLINK-25462
                 URL: https://issues.apache.org/jira/browse/FLINK-25462
             Project: Flink
          Issue Type: Improvement
          Components: API / DataStream
    Affects Versions: 1.14.2
            Reporter: Viktor Feklin


Currently there is no way to override partitioning for KeyedStream.

---------------------

Use case:

We use Kafka Source to read from topic where all records partitioned by some key.

We use same key for KeyedStream (all records with same key readed localy). 

We want to handle keyed events localy - because the load is already ballaced. 

Actualy events shuffle between nodes (because KeyedStream hardcoded with KeyGroupStreamPartitioner) and there is no way to stop this usless and expensive work (in our case). 

 



--
This message was sent by Atlassian Jira
(v8.20.1#820001)