[ 
https://issues.apache.org/jira/browse/KYLIN-1016?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14975521#comment-14975521
 ] 

Shaofeng SHI commented on KYLIN-1016:
-------------------------------------

Hi Dong, some comments here: how about the performance on a high cardinality 
column like > 10 million?

> Count distinct on any dimension should work even not a predefined measure
> -------------------------------------------------------------------------
>
>                 Key: KYLIN-1016
>                 URL: https://issues.apache.org/jira/browse/KYLIN-1016
>             Project: Kylin
>          Issue Type: Bug
>            Reporter: liyang
>            Assignee: liyang
>         Attachments: 
> 1.x-0001-KYLIN-1016-Count-distinct-on-any-dimension-should-wo.patch
>
>
> Currently count distinct on a dimension does not work:
> {"sql":"select  DATE'2015-07-18' , count(distinct country) as uniquecountry 
> from pc_session e INNER JOIN pc_cal c ON e.part_date = c.cal_dt WHERE 
> (part_date BETWEEN DATE'2015-07-18' AND DATE'2015-08-19') 
> ","offset":0,"limit":10,"acceptPartial":true,"project":"tracking"}
> "exception": "Can't find any realization. Please confirm with providers. SQL 
> digest: fact table DEFAULT.PC_SESSION,group by [],filter on 
> [DEFAULT.PC_SESSION.PART_DATE],with aggregates[FunctionDesc 
> [expression=COUNT_DISTINCT, parameter=ParameterDesc [type=column, 
> value=COUNTRY], returnType=null]].
> while executing SQL: "select  DATE'2015-07-18' , count(distinct country) as 
> uniquecountry from pc_session e INNER JOIN pc_cal c ON e.part_date = c.cal_dt 
> WHERE (part_date BETWEEN DATE'2015-07-18' AND DATE'2015-08-19') LIMIT 10""



--
This message was sent by Atlassian JIRA
(v6.3.4#6332)

Reply via email to