[
https://issues.apache.org/jira/browse/TEZ-2024?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14304463#comment-14304463
]
Hitesh Shah edited comment on TEZ-2024 at 2/4/15 1:38 AM:
----------------------------------------------------------
I think this issue is probably due to how we mingle both history and recovery.
From a recovery point of view, it might need a start before a finish. I think
the check was added for safety/correctness but you may be right that it might
not be needed after all. Are there known cases where an end is recorded even
though a start did not happen where the end state is not KILLED?
>From a history point of view, in some cases, a finish would help even if a
>start is not seen.
was (Author: hitesh):
I think this issue is probably due to how we mingle both history and recovery.
From a recovery point of view, it needs a start before a finish. From a history
point of view, in some cases, a finish would help even if a start is not seen.
> TaskFinishedEvent may not be logged in recovery
> ------------------------------------------------
>
> Key: TEZ-2024
> URL: https://issues.apache.org/jira/browse/TEZ-2024
> Project: Apache Tez
> Issue Type: Bug
> Reporter: Jeff Zhang
> Assignee: Jeff Zhang
> Attachments: TEZ-2024-1.patch
>
>
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
