[
https://issues.apache.org/jira/browse/YARN-4837?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Vinod Kumar Vavilapalli updated YARN-4837:
------------------------------------------
Attachment: YARN-4837-20160604.txt
Updated patch against the latest trunk.
[~kasha]
bq. RMAppAttemptImpl#shouldCountTowardsNodeBlacklisting: For
ContainerExitStatus.DISKS_FAILED, doesn't that mean at least one disk failed?
And, the NM can continue running with remaining disks right? Is the idea that
even if we schedule it to the same node, the NM wouldn't give the same local
directory? If yes, should we clarify the comment accordingly?
No, a container is marked with tContainerExitStatus.DISKS_FAILED means that
the node is already be marked unhealthy given that most of the disks failed.
So, no more containers will be scheduled on that node.
Edited the comment for more clarity to reflect the same.
> User facing aspects of 'AM blacklisting' feature need fixing
> ------------------------------------------------------------
>
> Key: YARN-4837
> URL: https://issues.apache.org/jira/browse/YARN-4837
> Project: Hadoop YARN
> Issue Type: Bug
> Reporter: Vinod Kumar Vavilapalli
> Assignee: Vinod Kumar Vavilapalli
> Priority: Critical
> Attachments: YARN-4837-20160515.txt, YARN-4837-20160520.1.txt,
> YARN-4837-20160520.txt, YARN-4837-20160527.txt, YARN-4837-20160604.txt
>
>
> Was reviewing the user-facing aspects that we are releasing as part of 2.8.0.
> Looking at the 'AM blacklisting feature', I see several things to be fixed
> before we release it in 2.8.0.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
---------------------------------------------------------------------
To unsubscribe, e-mail: yarn-issues-unsubscr...@hadoop.apache.org
For additional commands, e-mail: yarn-issues-h...@hadoop.apache.org
