You are viewing a plain text version of this content. The canonical link for it is here.
Posted to issues@kylin.apache.org by "Jacky Woo (JIRA)" <ji...@apache.org> on 2018/11/12 09:33:00 UTC

[jira] [Commented] (KYLIN-3613) Kylin with Standalone HBase Cluster (enabled kerberos) could not find the main cluster namespace at "Create HTable" step

    [ https://issues.apache.org/jira/browse/KYLIN-3613?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16683458#comment-16683458 ] 

Jacky Woo commented on KYLIN-3613:
----------------------------------

Hi,  I also met the same problem.  I add some logs  and found kylin use wrong hdfs when deploying coprocessor.  

But According the logic of  "HBaseConnection.java",  all values of "configThreadLocal" come from "newHBaseConfiguration" method, and "newHBaseConfiguration" method produces configuration with right hdfs.

What I wandering is why "HBaseConnection.getCurrentHBaseConfiguration" method get a configuration with wrong hdfs sometimes. 

> Kylin with Standalone HBase Cluster (enabled kerberos) could not find the main cluster namespace at  "Create HTable" step
> -------------------------------------------------------------------------------------------------------------------------
>
>                 Key: KYLIN-3613
>                 URL: https://issues.apache.org/jira/browse/KYLIN-3613
>             Project: Kylin
>          Issue Type: Bug
>          Components: Environment 
>    Affects Versions: v2.4.0, v2.4.1, v2.5.0
>            Reporter: powerinf
>            Assignee: WangSheng
>            Priority: Major
>             Fix For: v2.4.2, v2.5.1
>
>         Attachments: KYLIN-3613.patch, Kylin_standalone_hbase.log
>
>
> I deployed two hadoop cluster(also enabled kerberos ,with cross-realm trust) the main cluster and hbase cluster,Kylin Server can access both clusters using hdfs shell with fully qualifiered path ,can submit MR job to main cluster, and can use hive shell to access data warehouse
> on Kylin Server, the configurations of hadoop and hive points to main cluster,and can access hbase cluster using hbase shell.
> when I build the cube, at "Create HTable" step, it reported the error "java.net.UnknownHostException: ctyunbigdata Set hbase.table.sanity.checks to false 
> at conf or table descriptor if you want to bypass sanity checks",but after I restart Kylin serer , resume it can run normally,Why?
> more detail message on Kylin_standalone_hbase.log



--
This message was sent by Atlassian JIRA
(v7.6.3#76005)