[ 
https://issues.apache.org/jira/browse/YARN-1367?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
 ]

Anubhav Dhoot updated YARN-1367:
--------------------------------

    Attachment: YARN-1367.001.patch

Adds RESYNC_KEEPING_CONTAINERS as nodeaction that will cause NM to resync 
without killing containers. the older RESYNC is kept for handling the old RM as 
well as turning off this feature by making RM return RESYNC.
Refactored TestNodeManagerResync to do all test with both RESYNC apis while 
testing one preserves and one kills containers.

> After restart NM should resync with the RM without killing containers
> ---------------------------------------------------------------------
>
>                 Key: YARN-1367
>                 URL: https://issues.apache.org/jira/browse/YARN-1367
>             Project: Hadoop YARN
>          Issue Type: Sub-task
>          Components: resourcemanager
>            Reporter: Bikas Saha
>            Assignee: Anubhav Dhoot
>         Attachments: YARN-1367.001.patch, YARN-1367.prototype.patch
>
>
> After RM restart, the RM sends a resync response to NMs that heartbeat to it. 
>  Upon receiving the resync response, the NM kills all containers and 
> re-registers with the RM. The NM should be changed to not kill the container 
> and instead inform the RM about all currently running containers including 
> their allocations etc. After the re-register, the NM should send all pending 
> container completions to the RM as usual.



--
This message was sent by Atlassian JIRA
(v6.2#6252)

Reply via email to