[
https://issues.apache.org/jira/browse/MAPREDUCE-3359?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13144318#comment-13144318
]
Siddharth Seth commented on MAPREDUCE-3359:
-------------------------------------------
bq. I think key value pairs are too generic and can get very confusing to make
sense of.
Looking at JobConf and Job - they do already have lots of individual APIs
(along with allowing config via a conf object in Job). Yarn RM/NM have explicit
APIs only (App POV) - maybe it is better leaving it as is.
The list will just end up growing though - there's 2-3 log params which
eventually need to make it through to an NM, maybe other parameters later.
> Yarn clients / AM should be able to provide config options to the RM / NM
> -------------------------------------------------------------------------
>
> Key: MAPREDUCE-3359
> URL: https://issues.apache.org/jira/browse/MAPREDUCE-3359
> Project: Hadoop Map/Reduce
> Issue Type: Improvement
> Components: mrv2
> Affects Versions: 0.23.0
> Reporter: Siddharth Seth
> Assignee: Robert Joseph Evans
> Attachments: MR-3359.txt
>
>
> The RM and NM do not read a job's configuration. Clients / AMs should however
> be able to configure certain parameters for the RM/NM on a per app basis -
> like the Log Retention policy, token cancellation.
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators:
https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa
For more information on JIRA, see: http://www.atlassian.com/software/jira
