You are viewing a plain text version of this content. The canonical link for it is here.
Posted to issues@kylin.apache.org by "Peng Huang (Jira)" <ji...@apache.org> on 2019/09/19 02:11:00 UTC
[jira] [Updated] (KYLIN-4171) real-time query can't support
derivative field for condition
[ https://issues.apache.org/jira/browse/KYLIN-4171?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]
Peng Huang updated KYLIN-4171:
------------------------------
Affects Version/s: v3.0.0-alpha2
Description:
when i make a query like this
{code:java}
//sql
SELECT t.* FROM
( SELECT a.*,ROW_NUMBER() OVER(PARTITION BY data_id ORDER BY data_update_time desc) AS r
FROM kylin_stream_order a WHERE a.type = 'UPDATE') t
LIMIT 10
{code}
It's ok. I can get answer.
But when i make a query like this
{code:java}
//sql
SELECT t.* FROM
( SELECT a.*,ROW_NUMBER() OVER(PARTITION BY data_id ORDER BY data_update_time desc) AS r
FROM kylin_stream_order a WHERE a.type = 'UPDATE') t
WHERE t.r = 1
LIMIT 10
{code}
It's not ok. I get Nothing. However, same query for off-line cube, It's ok.
So, I think real-time query can't support derivative field for condition. Using `ROW_NUMBER` function to remove duplication is a common scenario. Please make real-time support it.
Summary: real-time query can't support derivative field for condition (was: real-time cube can)
> real-time query can't support derivative field for condition
> ------------------------------------------------------------
>
> Key: KYLIN-4171
> URL: https://issues.apache.org/jira/browse/KYLIN-4171
> Project: Kylin
> Issue Type: Bug
> Components: Real-time Streaming
> Affects Versions: v3.0.0-alpha2
> Reporter: Peng Huang
> Priority: Major
>
> when i make a query like this
>
> {code:java}
> //sql
> SELECT t.* FROM
> ( SELECT a.*,ROW_NUMBER() OVER(PARTITION BY data_id ORDER BY data_update_time desc) AS r
> FROM kylin_stream_order a WHERE a.type = 'UPDATE') t
> LIMIT 10
> {code}
> It's ok. I can get answer.
>
>
> But when i make a query like this
>
> {code:java}
> //sql
> SELECT t.* FROM
> ( SELECT a.*,ROW_NUMBER() OVER(PARTITION BY data_id ORDER BY data_update_time desc) AS r
> FROM kylin_stream_order a WHERE a.type = 'UPDATE') t
> WHERE t.r = 1
> LIMIT 10
> {code}
> It's not ok. I get Nothing. However, same query for off-line cube, It's ok.
>
>
> So, I think real-time query can't support derivative field for condition. Using `ROW_NUMBER` function to remove duplication is a common scenario. Please make real-time support it.
--
This message was sent by Atlassian Jira
(v8.3.4#803005)