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

Posted to issues@flink.apache.org by "Till Rohrmann (Jira)" <ji...@apache.org> on 2020/07/31 10:15:00 UTC

[jira] [Closed] (FLINK-11148) Rescaling operator regression in Flink 1.7

     [ https://issues.apache.org/jira/browse/FLINK-11148?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]

Till Rohrmann closed FLINK-11148.
---------------------------------
    Resolution: Duplicate

> Rescaling operator regression in Flink 1.7
> ------------------------------------------
>
>                 Key: FLINK-11148
>                 URL: https://issues.apache.org/jira/browse/FLINK-11148
>             Project: Flink
>          Issue Type: Bug
>          Components: Runtime / Coordination
>    Affects Versions: 1.7.0
>            Reporter: Truong Duc Kien
>            Priority: Major
>
> We have a job using 20 TaskManager with 3 slot each.
> Using Flink 1.4, when we rescale a data stream from 60 to 20, each TaskManager will only have one downstream slot, that receives the data from 3 upstream slots in the same TaskManager.
> Using Flink 1.7, this behaviour no longer hold true, multiple downstream slots are being assigned to the same TaskManager. This change is causing imbalance in our TaskManager load.



--
This message was sent by Atlassian Jira
(v8.3.4#803005)