[
https://issues.apache.org/jira/browse/FLINK-1003?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15364144#comment-15364144
]
ASF GitHub Bot commented on FLINK-1003:
---------------------------------------
Github user gyfora commented on the issue:
https://github.com/apache/flink/pull/2129
Thanks Stephan,
I agree that there is a lot to think about/improve when it comes to
scheduling and dynamic scaling. Should we add this to the Key groups design doc
or there is going to be a bigger one more general about different system
aspects of the dynamic scaling?
It might make sense to introduce some sort of a scheduling hint for when we
redeploy a job with higher parallelism so that we can exploit some state
checkpoint locality for quicker recovery, but this is another topic.
> Spread out scheduling strategy
> ------------------------------
>
> Key: FLINK-1003
> URL: https://issues.apache.org/jira/browse/FLINK-1003
> Project: Flink
> Issue Type: Improvement
> Reporter: Till Rohrmann
> Assignee: Gyula Fora
>
> Currently the Flink scheduler tries to fill one instance completely before
> the tasks are deployed to another instance. This is a good behaviour in
> multi-user and multi-job scenarios but it wastes resources if one wants to
> use the complete cluster. Therefore, another scheduling strategy where the
> load among the different instances is kept balanced might be useful. This
> spread out strategy will deploy the tasks such that the overall work is
> equally distributed.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
