[
https://issues.apache.org/jira/browse/TEZ-1216?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14094660#comment-14094660
]
Siddharth Seth commented on TEZ-1216:
-------------------------------------
The shutdown handling looks fine. My guess as to what is causing this is that
during serviceStop() - the moment the non-daemon thread that is keeping the AM
alive goes down, the shutdownhook can get invoked. This is assuming the
AMShutdownHandler is inheriting deamon properties from the thread that started
it.
bq. LOG.warn("Failed to tez scratch data dir", e);
Missing "delete" in the log statement.
Rest looks good to me.
> Clean up the staging directory when the application completes
> -------------------------------------------------------------
>
> Key: TEZ-1216
> URL: https://issues.apache.org/jira/browse/TEZ-1216
> Project: Apache Tez
> Issue Type: Sub-task
> Reporter: Mohammad Kamrul Islam
> Assignee: Hitesh Shah
> Priority: Critical
> Attachments: TEZ-1216.1.patch, TEZ-1216.2.patch, TEZ-1216.3.patch
>
>
> This JIRA is a part of TEZ-693. It includes the removal of stage directory
> when application's last attempt is done. In addition, it will check a config
> to determine whether to auto clean it or not.
--
This message was sent by Atlassian JIRA
(v6.2#6252)
