[ 
https://issues.apache.org/jira/browse/YARN-521?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13707792#comment-13707792
 ] 

Sandy Ryza commented on YARN-521:
---------------------------------

bq. It would not make sense to do so but if one does so then the method should 
return valid results, right?
Right

bq. The list looks good...
Thanks.  I think 99% of use cases will be covered by the first bullet point.  
The first three bullet points are actually covered by the last one, but (I 
hope) make it more explicit about when you don't need to worry.  So there 
aren't really places to comment them in the code, but I can try to make it 
clear that users shouldn't need to worry about grokking the last case if they 
fit into one of the earlier ones.

bq. Given this, how does one allow relaxing locality for a previously specific 
node request.
That's correct.  This is a consequence of the approach we decided on in 
YARN-392, in which the scheduler data structures can only accommodate one 
locality relaxation value at each priority.  Making this work would require 
modifying the scheduler data structures to account separately for 
locality-relaxed and non-locality-relaxed requests.  If we choose to do this in 
the future and remove the above restriction, I do not think it would be a 
backwards-incompatible change.

Uploading a patch with the additional requested tests and documentation.
                
> Augment AM - RM client module to be able to request containers only at 
> specific locations
> -----------------------------------------------------------------------------------------
>
>                 Key: YARN-521
>                 URL: https://issues.apache.org/jira/browse/YARN-521
>             Project: Hadoop YARN
>          Issue Type: Sub-task
>          Components: api
>    Affects Versions: 2.0.3-alpha
>            Reporter: Sandy Ryza
>            Assignee: Sandy Ryza
>         Attachments: YARN-521-1.patch, YARN-521-2.patch, YARN-521-2.patch, 
> YARN-521-3.patch, YARN-521-4.patch, YARN-521.patch
>
>
> When YARN-392 and YARN-398 are completed, it would be good for AMRMClient to 
> offer an easy way to access their functionality

--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira

Reply via email to