[ 
https://issues.apache.org/jira/browse/YARN-5124?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15316147#comment-15316147
 ] 

Arun Suresh edited comment on YARN-5124 at 6/6/16 6:00 AM:
-----------------------------------------------------------

[~kasha], thanks for the review..

I agree with 1-4 of your comments.. shall upload patch addressing them shortly..

w.r.t #5,
Given that {{ExecutionTypeRequest}} is a wrapper for the actual 
{{ExecutionType}}, the remoteRequestsTable primary key should comprise the 
ExecutionType, not the ExecutionTypeRequest. I was thinking, for the case where 
enforceExecutionType is false, we can possibly just match a returned Container 
with any entry in the remoteRequestsTable (where Priority, Location and 
Capability match, but the ExecutionType can be anything).
I intentionally left out the implementation, since currently, only Distributed 
Scheduling supports ExecutionTypes and it currently only supports 
enforceExecutionType = true. I would be happy to put at doc with a TODO there 
(and maybe open a JIRA) to fix it once we have the Scheduler fix that supports 
enforceExecutionType = false.
Thoughts ?


was (Author: asuresh):
[~kasha], thanks for the review..

I agree with 1-4 of your comments.. shall upload patch addressing them shortly..

w.r.t #5,
Given that {{ExecutionTypeRequest}} is a wrapper for the actual 
{{ExecutionType}}, the remoteRequestsTable primary key should comprise the 
ExecutionType, not the ExecutionTypeRequest. I was thinking, for the case where 
enforceExecutionType is false, we can possibly just match a returned Container 
with any entry in the remoteRequestsTable (where Priority, Location and 
Capability match, but the ExecutionType can be anything).
I intentionally left out the implementation, since currently, the only 
Distributed Scheduling supports ExecutionTypes and it currently only supports 
enforceExecutionType = true. I would be happy to put at doc with a TODO there 
(and maybe open a JIRA) to fix it once we have the Scheduler fix that supports 
enforceExecutionType = false.
Thoughts ?

> Modify AMRMClient to set the ExecutionType in the ResourceRequest
> -----------------------------------------------------------------
>
>                 Key: YARN-5124
>                 URL: https://issues.apache.org/jira/browse/YARN-5124
>             Project: Hadoop YARN
>          Issue Type: Sub-task
>            Reporter: Arun Suresh
>            Assignee: Arun Suresh
>         Attachments: YARN-5124.001.patch, YARN-5124.002.patch, 
> YARN-5124.003.patch, YARN-5124.004.patch, YARN-5124.005.patch, 
> YARN-5124.006.patch, YARN-5124.008.patch, 
> YARN-5124_YARN-5180_combined.007.patch, YARN-5124_YARN-5180_combined.008.patch
>
>
> Currently the {{ContainerRequest}} allows the AM to set the {{ExecutionType}} 
> in the AMRMClient, but it is not being set in the actual {{ResourceRequest}} 
> that is sent to the RM 



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

---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]

Reply via email to