[jira] [Commented] (KYLIN-3542) kylin cube导入数据到hive

["Shaofeng SHI (JIRA)" <ji...@apache.org>]

    [ https://issues.apache.org/jira/browse/KYLIN-3542?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16610338#comment-16610338 ] 

Shaofeng SHI commented on KYLIN-3542:
-------------------------------------

Kylin's Cube in HBase are encoded with dictionaries, so you couldn't expose them simply to Hive. 

 

If you do want to join historical data and real time, you can try to use "WITH " clause in one SQL, and join the results of these two clauses. The precondition is the result set of them should be small to be load into memory. For example:

 

https://stackoverflow.com/questions/12552288/sql-with-clause-example

> kylin cube导入数据到hive
> -------------------
>
>                 Key: KYLIN-3542
>                 URL: https://issues.apache.org/jira/browse/KYLIN-3542
>             Project: Kylin
>          Issue Type: Bug
>          Components: Integration
>    Affects Versions: v2.4.0
>            Reporter: pan
>            Priority: Blocker
>              Labels: features
>
> 听闻kylin 2.4已经支持kafka source与hive表关联，我现在有一个需求，我想把历史分析的结果与实时分析的结果向关联，也就是基于cube之前构建的数据和实时的数据关联分析，在kylin里面怎么实现呢？
> 我的思路是想办法吧把之前cube里面的数据映射到hive（hbase和hive整合），然后再接kafka数据源和hive表关联，但是在kylin里面segment有很多碎片，即使定时合并的话hbase表也是随时都在变，请问有没有办法在kylin里面实现hive和hbase整合呢？



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