hadoop-yarn-issues mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Íñigo Goiri (JIRA) <j...@apache.org>
Subject [jira] [Commented] (YARN-999) In case of long running tasks, reduce node resource should balloon out resource quickly by calling preemption API and suspending running task.
Date Tue, 12 Feb 2019 17:06:00 GMT

    [ https://issues.apache.org/jira/browse/YARN-999?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16766248#comment-16766248
] 

Íñigo Goiri commented on YARN-999:
----------------------------------

Thanks [~djp] for referring to YARN-2489.
I'll start working on a generic one and then we decide where to post it.

I think the idea would be for the RM to track the moment it got the change in resources and
once the timeout passes send {{ContainerPreemptEvent}}.
I see this is added in YARN-569 and used in a few places.

[~asuresh], [~kkaranasos], I remember you guys had work recently some preemption.
Do you guys know what would be a good JIRA to use as a reference for this?
Hopefully something that uses distinction of OPPORTUNISTIC containers and others.

> In case of long running tasks, reduce node resource should balloon out resource quickly
by calling preemption API and suspending running task. 
> -----------------------------------------------------------------------------------------------------------------------------------------------
>
>                 Key: YARN-999
>                 URL: https://issues.apache.org/jira/browse/YARN-999
>             Project: Hadoop YARN
>          Issue Type: Sub-task
>          Components: graceful, nodemanager, scheduler
>            Reporter: Junping Du
>            Priority: Major
>
> In current design and implementation, when we decrease resource on node to less than
resource consumption of current running tasks, tasks can still be running until the end. But
just no new task get assigned on this node (because AvailableResource < 0) until some tasks
are finished and AvailableResource > 0 again. This is good for most cases but in case of
long running task, it could be too slow for resource setting to actually work so preemption
could be hired here.



--
This message was sent by Atlassian JIRA
(v7.6.3#76005)

---------------------------------------------------------------------
To unsubscribe, e-mail: yarn-issues-unsubscribe@hadoop.apache.org
For additional commands, e-mail: yarn-issues-help@hadoop.apache.org


Mime
View raw message