You are viewing a plain text version of this content. The canonical link for it is here.

Posted to issues@kylin.apache.org by "ASF subversion and git services (Jira)" <ji...@apache.org> on 2019/12/05 06:02:00 UTC

[jira] [Commented] (KYLIN-4178) Job scheduler support safe mode

    [ https://issues.apache.org/jira/browse/KYLIN-4178?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16988496#comment-16988496 ] 

ASF subversion and git services commented on KYLIN-4178:
--------------------------------------------------------

Commit d55212e8bc5188fe1e9b3cb74818a4e4e49cff82 in kylin's branch refs/heads/document from Zhou Kang
[ https://gitbox.apache.org/repos/asf?p=kylin.git;h=d55212e ]

Add doc for KYLIN-4178


> Job scheduler support safe mode 
> --------------------------------
>
>                 Key: KYLIN-4178
>                 URL: https://issues.apache.org/jira/browse/KYLIN-4178
>             Project: Kylin
>          Issue Type: Improvement
>            Reporter: ZhouKang
>            Priority: Major
>             Fix For: v3.0.0
>
>         Attachments: image-2019-11-26-18-42-33-473.png
>
>
> Job scheduler should support safe mode in case of the HBase cluster change.
> In xiaomi, we want update the HBase cluster from hbase0.98 to HBase 2.0. The history data can be migrated previously, but the job has been submitted will keep running and write data to the old cluster. So we need a method to ensure that, job create htable in the old cluster will write data to the old cluster, and the job have not create htable should not be scheduled.
> So we need  job scheduler safe mode. Open safe mode before changing cluster config,  the  running jobs can run continuous, and the new job cannot be scheduled. 
> After all running job finished, we can change the cluster config to the new one,  and rest of job can be scheduled again.
>  



--
This message was sent by Atlassian Jira
(v8.3.4#803005)