[
https://issues.apache.org/jira/browse/YARN-4597?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15623825#comment-15623825
]
Jian He commented on YARN-4597:
-------------------------------
Few more comments
- maybe rename ContainerScheduler#runningContainers to scheduledContainers
- The ContainerLaunch#killedBeforeStart flag, looks like the exising flag
'shouldLaunchContainer' serves the same purpose, can we reuse that ? if so, the
container#isMarkedToKill is also not needed.
- NodeManager#containerScheduler variable not used, remove
- I think this comment is not addressed ? "In case we exceed the max-queue
length, we are killing the container directly instead of queueing the
container, in this case, we should not store the container as queued?"
> Add SCHEDULE to NM container lifecycle
> --------------------------------------
>
> Key: YARN-4597
> URL: https://issues.apache.org/jira/browse/YARN-4597
> Project: Hadoop YARN
> Issue Type: New Feature
> Components: nodemanager
> Reporter: Chris Douglas
> Assignee: Arun Suresh
> Labels: oct16-hard
> Attachments: YARN-4597.001.patch, YARN-4597.002.patch,
> YARN-4597.003.patch, YARN-4597.004.patch, YARN-4597.005.patch
>
>
> Currently, the NM immediately launches containers after resource
> localization. Several features could be more cleanly implemented if the NM
> included a separate stage for reserving resources.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
