[ 
https://issues.apache.org/jira/browse/YARN-2885?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15098525#comment-15098525
 ] 

Kishore Chaliparambil commented on YARN-2885:
---------------------------------------------

Hi [~asuresh],

I am reviewing the latest patch. I noticed that we assume that the last 
interceptor in the chain will be the LocalScheduler. This might break the model 
when we support YARN federation (YARN 3666). Federation interceptor will have 
to be the last interceptor since it abstracts the fact that there are multiple 
clusters from the application and clients. So I think instead of talking to the 
RM directly from the LocalScheduler, we can forward the request to the next 
interceptor in the chain. And until federation is implemented, we can have 
another interceptor implementation (e.g. DefaultRequestInterceptor ) that talks 
to the RM and use that as the last interceptor in the chain.

Thanks,
Kishore


> 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, 
> YARN-2885-yarn-2877.002.patch, YARN-2885-yarn-2877.full-2.patch, 
> YARN-2885-yarn-2877.full-3.patch, YARN-2885-yarn-2877.full.patch, 
> YARN-2885-yarn-2877.v4.patch, YARN-2885-yarn-2877.v5.patch, 
> YARN-2885_api_changes.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)

Reply via email to