[
https://issues.apache.org/jira/browse/YARN-3508?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14609096#comment-14609096
]
Varun Saxena commented on YARN-3508:
------------------------------------
BTW, as we are passing RMContext into {{SchedulingEditPolicy#init}}, we can do
away with PreemptableResourceScheduler as its parameter as well because we can
get scheduler instance from RMContext as well.
> Preemption processing occuring on the main RM dispatcher
> --------------------------------------------------------
>
> Key: YARN-3508
> URL: https://issues.apache.org/jira/browse/YARN-3508
> Project: Hadoop YARN
> Issue Type: Bug
> Components: resourcemanager, scheduler
> Affects Versions: 2.6.0
> Reporter: Jason Lowe
> Assignee: Varun Saxena
> Attachments: YARN-3508.002.patch, YARN-3508.01.patch,
> YARN-3508.03.patch, YARN-3508.04.patch, YARN-3508.05.patch
>
>
> We recently saw the RM for a large cluster lag far behind on the
> AsyncDispacher event queue. The AsyncDispatcher thread was consistently
> blocked on the highly-contended CapacityScheduler lock trying to dispatch
> preemption-related events for RMContainerPreemptEventDispatcher. Preemption
> processing should occur on the scheduler event dispatcher thread or a
> separate thread to avoid delaying the processing of other events in the
> primary dispatcher queue.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
