[
https://issues.apache.org/jira/browse/KYLIN-4862?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17258697#comment-17258697
]
JiangYang commented on KYLIN-4862:
----------------------------------
This issue is caused by two job engine scheduling the same one job.
>From logs, we can see
* at 08:07:42.106, kylin-all-0 released the lock
* at 08:07:42.132, kylin-job-0 acquired the lock
* at 09:56:37.774, kylin-all-0 tried to update job status because it was not
aware it had lost the lock. Then error occurs.
!Screen Shot 2020-11-11 at 4.21.38 PM.png!
!_thumb_2124727.png|width=1168,height=146,thumbnail!
The reason kylin-all-0 released the lock is because it was going through "Full
GC"
> Build Cube use two job engine, cause there is no valid state transfer
> from:ERROR to:SUCCEED
> -------------------------------------------------------------------------------------------
>
> Key: KYLIN-4862
> URL: https://issues.apache.org/jira/browse/KYLIN-4862
> Project: Kylin
> Issue Type: Improvement
> Components: Job Engine
> Reporter: JiangYang
> Assignee: JiangYang
> Priority: Major
>
--
This message was sent by Atlassian Jira
(v8.3.4#803005)
