[
https://issues.apache.org/jira/browse/YARN-1490?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13866264#comment-13866264
]
Jian He commented on YARN-1490:
-------------------------------
Uploaded a new patch
bq. Change it also transfer running containers too?
this is already there.
bq. Extend the test to explicitly validate that allocated/acquired/reserved are
killed ?
acquired to be killed test is already there. Add test for allocated container
to be killed.
bq. That TODO will be fixed in this ticket itself? Or a separate one?
fixed.
bq. We need to see if the remaining state needs to be transferred to. There is
some commented code.
I intentionally commented those, in case I forgot, I think for now those are
not needed yet.
bq. application.setShouldRecover(false);
Earlier I thought to reset the flag, but given we are always passing the
correct flag via AppAttemptRemovedSchedulerEvent, remove it.
bq. The following code-block can be in the constructor? if
(!appAttempt.submissionContext.getCleanContainersWhenFail()) {
bq. Similarly the check for unmanaged AM can also be in the constructor?
This flag should apply only when the attempt is failed, not for killed/
finished. right? Added test cases for this.
bq. Can you file a ticket for the broken stuff in unmanaged-AM?
done. YARN-1577
> RM should optionally not kill all containers when an ApplicationMaster exits
> ----------------------------------------------------------------------------
>
> Key: YARN-1490
> URL: https://issues.apache.org/jira/browse/YARN-1490
> Project: Hadoop YARN
> Issue Type: Sub-task
> Reporter: Vinod Kumar Vavilapalli
> Assignee: Jian He
> Attachments: YARN-1490.1.patch, YARN-1490.2.patch, YARN-1490.3.patch,
> YARN-1490.4.patch
>
>
> This is needed to enable work-preserving AM restart. Some apps can chose to
> reconnect with old running containers, some may not want to. This should be
> an option.
--
This message was sent by Atlassian JIRA
(v6.1.5#6160)
