[
https://issues.apache.org/jira/browse/KYLIN-4007?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16843878#comment-16843878
]
XiaoXiang Yu edited comment on KYLIN-4007 at 5/20/19 11:28 AM:
---------------------------------------------------------------
Looks like this issue is related to
https://issues.apache.org/jira/browse/KYLIN-3633. But it should be fixed at
v2.5.1, maybe it caused by other exception. But without more log/ stack trace,
it difficult to detect in which way should we repair.
was (Author: hit_lacus):
Looks like this issue is related to
https://issues.apache.org/jira/browse/KYLIN-3633.
> Kylin concurrently builds a global dictionary for too long
> ----------------------------------------------------------
>
> Key: KYLIN-4007
> URL: https://issues.apache.org/jira/browse/KYLIN-4007
> Project: Kylin
> Issue Type: Bug
> Components: Job Engine
> Affects Versions: v2.6.1
> Reporter: mouxingya
> Assignee: XiaoXiang Yu
> Priority: Minor
> Attachments: WechatIMG20.png
>
>
> When a dimension (user_id) is accurately calculated at the same time, the
> original data is in the order of millions, and the rowkey has the encoding
> type fixed_lengrh and length 100. It took a few hours or even failed in the
> build dimension dictionary step. After investigation, it was found that there
> are concurrent processes that want to obtain the lock corresponding to the
> cube / dict path, resulting in a deadlock.
--
This message was sent by Atlassian JIRA
(v7.6.3#76005)
