[
https://issues.apache.org/jira/browse/YARN-2885?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15038768#comment-15038768
]
Wangda Tan commented on YARN-2885:
----------------------------------
Hi [~asuresh],
Thanks for reply,
bq. What we were aiming for is to not send any Queueable resource reqs to the
RM...
After thinking, RM could directly support queue-able container allocation.
Since queue-able/guaranteed executionType is part of user-facing API, so
scheduler can consider to allocate queue-able container or not. LocalRM is a
way to allocate queue-able containers. But please make sure that there's no
assumption (hardcoded logic) that queue-able container can be only allocated by
LocalRM?
bq. I totally agree that the AM should not be bothered with this.. But if you
notice, It is actually not set by the AM, it set by the
DistSchedulerReqeustInterceptor when it proxies the AM calls...
Since you planned to have a LocalRM coordinator, I would prefer to add a
separated Distributed Scheduler Coordinator service and protocols.
Other comments are make sense to me.
> Create AMRMProxy request interceptor for distributed scheduling decisions for
> queueable containers
> --------------------------------------------------------------------------------------------------
>
> Key: YARN-2885
> URL: https://issues.apache.org/jira/browse/YARN-2885
> Project: Hadoop YARN
> Issue Type: Sub-task
> Components: nodemanager, resourcemanager
> Reporter: Konstantinos Karanasos
> Assignee: Arun Suresh
> Attachments: YARN-2885-yarn-2877.001.patch
>
>
> We propose to add a Local ResourceManager (LocalRM) to the NM in order to
> support distributed scheduling decisions.
> Architecturally we leverage the RMProxy, introduced in YARN-2884.
> The LocalRM makes distributed decisions for queuable containers requests.
> Guaranteed-start requests are still handled by the central RM.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
