[
https://issues.apache.org/jira/browse/YARN-2882?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14221288#comment-14221288
]
Konstantinos Karanasos commented on YARN-2882:
----------------------------------------------
[~Sujeet Varakhedi] The scope of pre-emption/killing is not only within an
application. Whenever a guaranteed-start task arrives in an NM that cannot
accommodate its execution due to running queueable tasks, it is allowed to
pre-empt/kill one or more of those, even if they belong to another application.
Clearly there can be policies that decide which of the running queueable tasks
to pre-empt/kill (and one of them could be to avoid pre-empting/killing a task
of another application, if there is a good reason for that).
> Introducing container types
> ---------------------------
>
> Key: YARN-2882
> URL: https://issues.apache.org/jira/browse/YARN-2882
> Project: Hadoop YARN
> Issue Type: Sub-task
> Components: nodemanager, resourcemanager
> Reporter: Konstantinos Karanasos
>
> This JIRA introduces the notion of container types.
> We propose two initial types of containers: guaranteed-start and queueable
> containers.
> Guaranteed-start are the existing containers, which are allocated by the
> central RM and are instantaneously started, once allocated.
> Queueable is a new type of container, which allows containers to be queued in
> the NM, thus their execution may be arbitrarily delayed.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
