[jira] [Updated] (YARN-3453) Fair Scheduler : Parts of preemption logic uses DefaultResourceCalculator even in DRF mode causing thrashing

["Arun Suresh (JIRA)" <ji...@apache.org>]

     [ https://issues.apache.org/jira/browse/YARN-3453?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]

Arun Suresh updated YARN-3453:
------------------------------
    Attachment: YARN-3453.1.patch

[~peng.zhang], [~ashwinshankar77],
Thank you reporting this.. and the associated discussion

I vote that we :
# fix the {{isStarved()}} method to use the correct Calculator
# fix the {{resToPreempt()}} method to use componentWiseMin for the target... but defer using the {{targetRatio}}, since it is probably an optimization and can be addressed in a future JIRA

I have attached a preliminary patch that does this..
Will upload one with test cases shortly


> Fair Scheduler : Parts of preemption logic uses DefaultResourceCalculator even in DRF mode causing thrashing
> ------------------------------------------------------------------------------------------------------------
>
>                 Key: YARN-3453
>                 URL: https://issues.apache.org/jira/browse/YARN-3453
>             Project: Hadoop YARN
>          Issue Type: Bug
>          Components: fairscheduler
>    Affects Versions: 2.6.0
>            Reporter: Ashwin Shankar
>            Assignee: Arun Suresh
>         Attachments: YARN-3453.1.patch
>
>
> There are two places in preemption code flow where DefaultResourceCalculator is used, even in DRF mode.
> Which basically results in more resources getting preempted than needed, and those extra preempted containers aren’t even getting to the “starved” queue since scheduling logic is based on DRF's Calculator.
> Following are the two places :
> 1. {code:title=FSLeafQueue.java|borderStyle=solid}
> private boolean isStarved(Resource share)
> {code}
> A queue shouldn’t be marked as “starved” if the dominant resource usage
> is >=  fair/minshare.
> 2. {code:title=FairScheduler.java|borderStyle=solid}
> protected Resource resToPreempt(FSLeafQueue sched, long curTime)
> {code}
> --------------------------------------------------------------
> One more thing that I believe needs to change in DRF mode is : during a preemption round,if preempting a few containers results in satisfying needs of a resource type, then we should exit that preemption round, since the containers that we just preempted should bring the dominant resource usage to min/fair share.



--
This message was sent by Atlassian JIRA
(v6.3.4#6332)