[
https://issues.apache.org/jira/browse/KYLIN-3837?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17116542#comment-17116542
]
langdamao commented on KYLIN-3837:
----------------------------------
I got the same problem few days ago. The reason is hadoop3.0.0 use port 9820,
but use port 8020 for other version. detail
in[HDFS-12990|https://issues.apache.org/jira/browse/HDFS-12990?jql=text%20~%20%229820%22].
To fix the problem try to use jar with hadoop other version.
Meanwhile，we found that the hadoop jar we use to connect HDFS in a Job is in
hbase-lib, we rm /hbase/lib/hadoop-*-3.0.0.jar，and use system hdfs lib to
sloved the problem. The hadoop jar using may be for different environment，but
not only in hdfs-lib.
> hadoop3.0 mapreduce task can not run successfully
> -------------------------------------------------
>
> Key: KYLIN-3837
> URL: https://issues.apache.org/jira/browse/KYLIN-3837
> Project: Kylin
> Issue Type: Bug
> Components: Job Engine
> Affects Versions: v2.5.1
> Reporter: 梁龙
> Priority: Major
> Attachments: image-2019-03-04-09-59-28-862.png,
> image-2019-03-04-10-00-33-056.png, image-2019-03-04-10-03-04-021.png
>
>
> At the kylin_streaming_cube demo，The first step is to make a mistake.
> !image-2019-03-04-09-59-28-862.png!
>
> Output ：
> !image-2019-03-04-10-00-33-056.png!
>
> and the mapreduce error:
> !image-2019-03-04-10-03-04-021.png!
>
>
--
This message was sent by Atlassian Jira
(v8.3.4#803005)
