[jira] [Commented] (KYLIN-239) Build Cube with sampling/partial data

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

    [ https://issues.apache.org/jira/browse/KYLIN-239?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15931430#comment-15931430 ] 

liyang commented on KYLIN-239:
------------------------------

This is somewhat the same as incrementally build a small piece of data from source.

Think it is already supported.

> Build Cube with sampling/partial data
> -------------------------------------
>
>                 Key: KYLIN-239
>                 URL: https://issues.apache.org/jira/browse/KYLIN-239
>             Project: Kylin
>          Issue Type: Wish
>          Components: General
>            Reporter: Luke Han
>            Assignee: liyang
>            Priority: Minor
>              Labels: github-import
>
> Debug and test cube with full data is not cost-effective: long time to build cube, heavy MR jobs to process and big storage of final HBase. 
> Extract sample data from Hive and run sample cube based on it will bring more benefits:
> 1. Rapid iterate cube design when enable new data model in Kylin
> 2. Easy to predicate final storage, query latency
> 3. Fast enable end to end solution
> 4. Quick duplicate on Dev env when there's issue on Prod
> ---------------- Imported from GitHub ----------------
> Url: https://github.com/KylinOLAP/Kylin/issues/266
> Created by: [lukehan|https://github.com/lukehan]
> Labels: 
> Milestone: Backlog
> Created at: Fri Dec 26 11:21:16 CST 2014
> State: open



--
This message was sent by Atlassian JIRA
(v6.3.15#6346)