[jira] [Resolved] (YARN-4134) FairScheduler preemption stops at queue level that all child queues are not over their fairshare

["Karthik Kambatla (JIRA)" <ji...@apache.org>]

     [ https://issues.apache.org/jira/browse/YARN-4134?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]

Karthik Kambatla resolved YARN-4134.
------------------------------------
    Resolution: Duplicate

[~xinxianyin] - thanks for reporting and working on this. As you say, this duplicates YARN-3405. I am looking to some comprehensive preemption changes - let us discuss your solution approach on YARN-3405 and see how any solution fits into the overall picture at YARN-4752. Thanks. 

> FairScheduler preemption stops at queue level that all child queues are not over their fairshare
> ------------------------------------------------------------------------------------------------
>
>                 Key: YARN-4134
>                 URL: https://issues.apache.org/jira/browse/YARN-4134
>             Project: Hadoop YARN
>          Issue Type: Sub-task
>          Components: fairscheduler
>            Reporter: Xianyin Xin
>            Assignee: Xianyin Xin
>         Attachments: YARN-4134.001.patch, YARN-4134.002.patch, YARN-4134.003.patch
>
>
> Now FairScheudler uses a choose-a-candidate method to select a container from leaf queues that to be preempted, in {{FSParentQueue.preemptContainer()}},
> {code}
>     readLock.lock();
>     try {
>       for (FSQueue queue : childQueues) {
>         if (candidateQueue == null ||
>             comparator.compare(queue, candidateQueue) > 0) {
>           candidateQueue = queue;
>         }
>       }
>     } finally {
>       readLock.unlock();
>     }
>     // Let the selected queue choose which of its container to preempt
>     if (candidateQueue != null) {
>       toBePreempted = candidateQueue.preemptContainer();
>     }
> {code}
> a candidate child queue is selected. However, if the queue's usage isn't over it's fairshare, preemption will not happen:
> {code}
>     if (!preemptContainerPreCheck()) {
>       return toBePreempted;
>     }
> {code}
>  A scenario:
> {code}
>             root
>            /    \
>       queue1   queue2
>                /    \
>           queue2.3, (  queue2.4  )
> {code}
> suppose there're 8 containers, and queues at any level have the same weight. queue1 takes 4 and queue2.3 takes 4, so both queue1 and queue2 are at their fairshare. Now we submit an app in queue2.4 with 4 containers needs, it should preempt 2 from queue2.3, but the candidate-containers selection procedure will stop at queue1, so none of the containers will be preempted.



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