[
https://issues.apache.org/jira/browse/YARN-8137?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16440103#comment-16440103
]
Giovanni Matteo Fumarola edited comment on YARN-8137 at 4/16/18 10:20 PM:
--------------------------------------------------------------------------
Thanks [~abmodi] for the patch.
Few comments:
1) You have this instruction {{int threadPoolSize = Math.max(poolSize, 10);}}
instead of 10 you should have {{SLSConfiguration.RUNNER_POOL_SIZE_DEFAULT;}}
{{2) {{executorService.awaitTermination(1, TimeUnit.HOURS);}}}} I think the
awaiting time is too long. How did you come up with this value?
was (Author: giovanni.fumarola):
Thanks [~abmodi] for the patch.
Few comments:
1) You have this instruction {{int threadPoolSize = Math.max(poolSize, 10);}}
instead of 10 you should have {{SLSConfiguration.RUNNER_POOL_SIZE_DEFAULT;}}
{{2) }}{{executorService.awaitTermination(1, TimeUnit.HOURS);}} I think the
awaiting time is too long. How did you come up with this value?
> Parallelize node addition in SLS
> --------------------------------
>
> Key: YARN-8137
> URL: https://issues.apache.org/jira/browse/YARN-8137
> Project: Hadoop YARN
> Issue Type: Sub-task
> Reporter: Abhishek Modi
> Assignee: Abhishek Modi
> Priority: Major
> Attachments: YARN-8137.001.patch
>
>
> Right now, nodes are added sequentially and it can take a long time if there
> are large number of nodes. With this change nodes will be added in parallel
> and thus reduce the node addition time.
--
This message was sent by Atlassian JIRA
(v7.6.3#76005)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
