hadoop-yarn-issues mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Arun Suresh (JIRA)" <j...@apache.org>
Subject [jira] [Commented] (YARN-5292) Support for PAUSED container state
Date Fri, 24 Jun 2016 15:35:16 GMT

    [ https://issues.apache.org/jira/browse/YARN-5292?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15348434#comment-15348434

Arun Suresh commented on YARN-5292:

Appreciate the feedback [~jlowe]..
We were initially thinking of FREEZE / THAW. Guessing that would be more indicative.

With regard to storing state, any thoughts of if we should expose an NM service / API to containers,
which can be used by the container to store/load state prior to free/thawing, or should the
NM just assume the Container is just notified of the fact that it will be frozen/thawed and
it handles state persistence.

Also, YARN-4876 allows AM's to explicitly stop / start a containers while retaining the allocation
for the container. We should clarify the difference between a stopped (but can be re-started)
containers and a frozen container. One difference could probably be that container stopping
is initialed by the AM and am guessing container freezing would be initiated by the NM container

> Support for PAUSED container state
> ----------------------------------
>                 Key: YARN-5292
>                 URL: https://issues.apache.org/jira/browse/YARN-5292
>             Project: Hadoop YARN
>          Issue Type: New Feature
>            Reporter: Hitesh Sharma
> YARN-2877 introduced OPPORTUNISTIC containers, and YARN-5216 proposes to add capability
to customize how OPPORTUNISTIC containers get preempted.
> In this JIRA we propose introducing a PAUSED container state.
> When a running container gets preempted, it enters the PAUSED state, where it remains
until resources get freed up on the node then the preempted container can resume to the running
> One scenario where this capability is useful is work preservation. How preemption is
done, and whether the container supports it, is implementation specific.
> For instance, if the container is a virtual machine, then preempt would pause the VM
and resume would restore it back to the running state.
> If the container doesn't support preemption, then preempt would default to killing the

This message was sent by Atlassian JIRA

To unsubscribe, e-mail: yarn-issues-unsubscribe@hadoop.apache.org
For additional commands, e-mail: yarn-issues-help@hadoop.apache.org

View raw message