[
https://issues.apache.org/jira/browse/FLINK-12122?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16845683#comment-16845683
]
Nico Kruber commented on FLINK-12122:
-------------------------------------
Since every subtask gets its own thread (not every slot / slot sharing group),
you don't need to split into slot sharing groups in order to have concurrent
processing. You may just need to break an operator chain.
What I was saying: if you have a parallelism 50 job and only 50 slots in total,
you'll get your even distribution - as a workaround at least (I'm not saying
anything against offering the ability for users to control this behaviour).
Apparently, you have more slots than your desired parallelism.
> Spread out tasks evenly across all available registered TaskManagers
> --------------------------------------------------------------------
>
> Key: FLINK-12122
> URL: https://issues.apache.org/jira/browse/FLINK-12122
> Project: Flink
> Issue Type: Sub-task
> Components: Runtime / Coordination
> Affects Versions: 1.6.4, 1.7.2, 1.8.0
> Reporter: Till Rohrmann
> Assignee: Till Rohrmann
> Priority: Major
> Fix For: 1.7.3, 1.9.0, 1.8.1
>
> Attachments: image-2019-05-21-12-28-29-538.png,
> image-2019-05-21-13-02-50-251.png
>
>
> With Flip-6, we changed the default behaviour how slots are assigned to
> {{TaskManages}}. Instead of evenly spreading it out over all registered
> {{TaskManagers}}, we randomly pick slots from {{TaskManagers}} with a
> tendency to first fill up a TM before using another one. This is a regression
> wrt the pre Flip-6 code.
> I suggest to change the behaviour so that we try to evenly distribute slots
> across all available {{TaskManagers}} by considering how many of their slots
> are already allocated.
--
This message was sent by Atlassian JIRA
(v7.6.3#76005)
