[
https://issues.apache.org/jira/browse/TEZ-1827?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14354098#comment-14354098
]
Bikas Saha edited comment on TEZ-1827 at 3/10/15 1:52 AM:
----------------------------------------------------------
Would this create a race condition that allows the cluster to shutdown because
the test finished but before the AMs could unregister from the RM. This would
leave hung App master processes after the test. Perhaps we should rename the
config to kill timeout and use it to trigger the kill but still wait after that
for the apps to finish. If the apps are not finishing after that then they
might be revealing some other legitimate issue. The kill timeout config could
be more aggressive.
+1 for timeline cluster inheriting from tez cluster.
was (Author: bikassaha):
Would this create a race condition that allows the cluster to shutdown because
the test finished but before the AMs could unregister from the RM. This would
leave hung App master processes after the test. Perhaps we should rename the
config to kill timeout and use it to trigger the kill but still wait after that
for the apps to finish. If the apps are not finishing after that then they
might be revealing some other legitimate issue.
> MiniTezCluster takes 10 minutes to shut down
> --------------------------------------------
>
> Key: TEZ-1827
> URL: https://issues.apache.org/jira/browse/TEZ-1827
> Project: Apache Tez
> Issue Type: Bug
> Affects Versions: 0.5.2
> Reporter: Ashutosh Chauhan
> Assignee: Hitesh Shah
> Priority: Minor
> Attachments: TEZ-1827.1.patch
>
>
> Always reproducible
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
