Hi, @zhongxianglai Thanks for your response.

This is what we're doing.

Hive on HDP1
    |
    |----------------Kylin Primary    -------------- HBase on HDP1
    |
    |----------------Kylin Secondary ------------- HBase on HDP2

Everything works fine on primary. The issue is on the Secondary Kylin.

I've tried your method on 
https://kylin.apache.org/blog/2016/06/10/standalone-hbase-cluster/
But it seems doesn't work. (I'm using kylin3.1.2, HDP3.1, HBase 2.0, but the 
doc was created 5 years ago)

On secondary Kylin, from the log, I can see 

“””
……
CREATE EXTERNAL TABLE IF NOT EXISTS 
kylin_intermediate_test_cart_add_cube_0ed9de6e_80eb_1d90_dd3f_9ef85ddd8941
(
\`CART_ADD_USER_KEY\` string
)
STORED AS SEQUENCEFILE
LOCATION 
'hdfs://HDP2_NN/tmp/aipl/kylin/aipl-kylin_metadata/kylin-6db3c2d5-49f0-dee4-adee-45f689d05460/kylin_intermediate_test_cart_add_cube_0ed9de6e_80eb_1d90_dd3f_9ef85ddd8941'
;
……
”””
Kylin is connecting to hive server on HDP1
But the LOCATION in the sql is pointing to NN on HDP2
So the error is 
Error: Error while compiling statement: FAILED: SemanticException 
java.lang.IllegalArgumentException: java.net.UnknownHostException: HDP2_NN 
(state=42000,code=40000)

I also tried merging hdfs-site.xml on HDP1 and HDP2 and put it in the client 
side (I couldn't change anything on server side), but finally failed.

So, is there any other solution?

Thank you!

Best regards,
____________________________
Lingsheng Meng | Forest | CNTD




On 2022/1/11, 8:20 PM, "zhongxianglai" <[email protected]> wrote:

    [EXTERNAL] This message comes from an external organization.

    退订

    -----邮件原件-----
    发件人: [email protected] [mailto:[email protected]] 代表 Xiaoxiang Yu
    发送时间: 2022年1月11日 19:07
    收件人: [email protected]; [email protected]
    主题: Re: How to install Kylin3 on independent Hbase2

    From blog 
"https://kylin.apache.org/blog/2016/06/10/standalone-hbase-cluster/",  you can 
see the sentence "Make sure Kylin Server can submit MR job to main cluster, and 
can use hive shell to access data warehouse, make sure the configurations of 
hadoop and hive points to main cluster " .


    And, you can check another doc which only for Kylin 4, "Copy the 
hive-site.xml from build cluster to query cluster into right directory 
(typically, hive configuration should located at : /etc/hive/conf). So the 
query can load Hive metadata from build cluster properly."  show the same 
things. (Link : 
https://cwiki.apache.org/confluence/display/KYLIN/Read-Write+Separation+Deployment+for+Kylin+4.0)


    So I think maybe you forget to copy right hive-site.xml to HBase cluster?




    --

    Best wishes to you !
    From :Xiaoxiang Yu

Reply via email to