[ 
https://issues.apache.org/jira/browse/YARN-562?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13638568#comment-13638568
 ] 

Jian He commented on YARN-562:
------------------------------

bq. Can we signal an object here so that we can be notified here that we are 
done launching stale containers
The purpose I'm sleeping inside getNodeStatusAndUpdateContainersInContext() is 
not for synchronize purpose. I had actually already join the thread in 
overridden rebootNodeStatusUpdater(). The goal here is to simulate that 
"while(!containers.isEmpty())" is taking longer. Because its very likely that 
it goes out of loop very quickly(only 1 running container) but the test thread 
is still launching a bunch of containers. We want to test the wait loop will 
not go indefinitely, while someone else keeps launching new containers

bq. We could start the launcher thread in the test code after sending the 
RESYNC event
I dont think we can start here, because its possible,though unlikely,the test 
thread has finished, but we have not reached inside the cleanupContainers 
method at all 
                
> NM should reject containers allocated by previous RM
> ----------------------------------------------------
>
>                 Key: YARN-562
>                 URL: https://issues.apache.org/jira/browse/YARN-562
>             Project: Hadoop YARN
>          Issue Type: Sub-task
>          Components: resourcemanager
>            Reporter: Jian He
>            Assignee: Jian He
>         Attachments: YARN-562.1.patch, YARN-562.2.patch, YARN-562.3.patch, 
> YARN-562.4.patch, YARN-562.5.patch, YARN-562.6.patch, YARN-562.7.patch
>
>
> Its possible that after RM shutdown, before AM goes down,AM still call 
> startContainer on NM with containers allocated by previous RM. When RM comes 
> back, NM doesn't know whether this container launch request comes from 
> previous RM or the current RM. we should reject containers allocated by 
> previous RM 

--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira

Reply via email to