You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@kylin.apache.org by 仇同心 <qi...@jd.com> on 2016/07/11 08:29:56 UTC
答复: cube查询问题
少峰你好：
Kylin的下个版本对于这些bug都会做修复吗？下个版本何时会发布？
谢谢！
-----邮件原件-----
发件人: ShaoFeng Shi [mailto:shaofengshi@apache.org]
发送时间: 2016年7月11日 16:25
收件人: dev; nidongdong1987
主题: Re: cube查询问题
There is a bug for "union all":
https://issues.apache.org/jira/browse/KYLIN-1833
Yu Feng provided a patch but may haven't been merged.
2016-07-11 15:59 GMT+08:00 nidongdong1987 <ni...@126.com>:
> dev：
>
> 1. 软件环境：
>
> kylin：1.5.2-cdh5.7.1
> hadoop：2.6.0-cdh5.7.1
> hive：1.1.0-cdh5.7.1
> hbase：1.2.0-cdh5.7.1
> JDK: 1.7.0_67
>
> 2. 问题描述：
>
> 在Kylin Web GUI用union all合并两个cube结果时，返回结果不正确。哪位知道是什么原因导致
> 例：
> 1）分别查询两张已建cube的表
>
> 查询sql：
> 查询sql：
> select dim01,
> select dim01,
> sum(val_cnt) as val_cnt1,
> sum(val_cnt) as val_cnt1,
> 1 as flag
> 2 as flag
> from td_idx_1000100001_1
> from td_idx_1000100002_1
> group by dim01；
> group by dim01；
>
> 返回结果：
> 返回结果：
> DIM01 VAL_CNT1 FLAG
> DIM01 VAL_CNT1 FLAG
> A0001 40031.37 1
> A0001 40207.42 2
> A0002 41613.63 1
> A0002 41698.91 2
> A0003 42260.19 1
> A0003 42390.84 2
>
> 2）用union all进行合并查询
>
> 查询sql：
> select dim01,
> sum(val_cnt) as val_cnt1,
> 1 as flag
> from td_idx_1000100001_1
> group by dim01
> union all
> select dim01,
> sum(val_cnt) as val_cnt1,
> 2 as flag
> from td_idx_1000100002_1
> group by dim01
>
> 预期返回结果： 实际返回结果：
> DIM01 VAL_CNT1 FLAG DIM01 VAL_CNT1 FLAG
> A0001 40031.37 1 A0001 40207.42 1
> A0002 41613.63 1 A0002 41698.91 1
> A0003 42260.19 1 A0003 42390.84 1
> A0001 40207.42 2 A0001 40207.42 2
> A0002 41698.91 2 A0002 41698.91 2
> A0003 42390.84 2 A0003 42390.84 2
>
> 发现FLAG =1 时的 VAL_CNT1 值不正确，查询的值恰好是最后一个union all表的值。
>
> 3）多表union all 及两张表left join亦是如此
>
>
> 相关日志：
> SQL: select dim01,
> sum(val_cnt) as val_cnt1,
> 1 as flag
> from td_idx_1000100001_1
> group by dim01
> union all
> select dim01,
> sum(val_cnt) as val_cnt1,
> 2 as flag
> from td_idx_1000100002_1
> group by dim01
> order by flag,dim01
> User: ADMIN
> Success: true
> Duration: 0.069
> Project: asiainfo_kylin
> Realization Names: [cube_1000100002_1]
> Cuboid Ids: [16]
> Total scan count: 6
> Result row count: 6
> Accept Partial: true
> Is Partial Result: false
> Hit Exception Cache: false
> Storage cache used: false
> Message: null
>
>
>
>
>
>
> nidongdong1987
--
Best regards,
Shaofeng Shi
Re: 答复: cube查询问题
Posted by ShaoFeng Shi <sh...@gmail.com>.
hi，we have kick off the 1.5.3 release process in last week, will invite the community to do testing soon. The plan is to finish it in this month.
For this "union" bug, since it wasn't marked as "resolved" when kick off the release, it may have to defer to next version.
Regards，
Shaofeng Shi
shaofengshi@gmail.com
From Outlook Mobile
On Mon, Jul 11, 2016 at 4:30 PM +0800, "仇同心" <qi...@jd.com> wrote:
少峰你好：
Kylin的下个版本对于这些bug都会做修复吗？下个版本何时会发布？
谢谢！
-----邮件原件-----
发件人: ShaoFeng Shi [mailto:shaofengshi@apache.org]
发送时间: 2016年7月11日 16:25
收件人: dev; nidongdong1987
主题: Re: cube查询问题
There is a bug for "union all":
https://issues.apache.org/jira/browse/KYLIN-1833
Yu Feng provided a patch but may haven't been merged.
2016-07-11 15:59 GMT+08:00 nidongdong1987 :
> dev：
>
> 1. 软件环境：
>
> kylin：1.5.2-cdh5.7.1
> hadoop：2.6.0-cdh5.7.1
> hive：1.1.0-cdh5.7.1
> hbase：1.2.0-cdh5.7.1
> JDK: 1.7.0_67
>
> 2. 问题描述：
>
> 在Kylin Web GUI用union all合并两个cube结果时，返回结果不正确。哪位知道是什么原因导致
> 例：
> 1）分别查询两张已建cube的表
>
> 查询sql：
> 查询sql：
> select dim01,
> select dim01,
> sum(val_cnt) as val_cnt1,
> sum(val_cnt) as val_cnt1,
> 1 as flag
> 2 as flag
> from td_idx_1000100001_1
> from td_idx_1000100002_1
> group by dim01；
> group by dim01；
>
> 返回结果：
> 返回结果：
> DIM01 VAL_CNT1 FLAG
> DIM01 VAL_CNT1 FLAG
> A0001 40031.37 1
> A0001 40207.42 2
> A0002 41613.63 1
> A0002 41698.91 2
> A0003 42260.19 1
> A0003 42390.84 2
>
> 2）用union all进行合并查询
>
> 查询sql：
> select dim01,
> sum(val_cnt) as val_cnt1,
> 1 as flag
> from td_idx_1000100001_1
> group by dim01
> union all
> select dim01,
> sum(val_cnt) as val_cnt1,
> 2 as flag
> from td_idx_1000100002_1
> group by dim01
>
> 预期返回结果： 实际返回结果：
> DIM01 VAL_CNT1 FLAG DIM01 VAL_CNT1 FLAG
> A0001 40031.37 1 A0001 40207.42 1
> A0002 41613.63 1 A0002 41698.91 1
> A0003 42260.19 1 A0003 42390.84 1
> A0001 40207.42 2 A0001 40207.42 2
> A0002 41698.91 2 A0002 41698.91 2
> A0003 42390.84 2 A0003 42390.84 2
>
> 发现FLAG =1 时的 VAL_CNT1 值不正确，查询的值恰好是最后一个union all表的值。
>
> 3）多表union all 及两张表left join亦是如此
>
>
> 相关日志：
> SQL: select dim01,
> sum(val_cnt) as val_cnt1,
> 1 as flag
> from td_idx_1000100001_1
> group by dim01
> union all
> select dim01,
> sum(val_cnt) as val_cnt1,
> 2 as flag
> from td_idx_1000100002_1
> group by dim01
> order by flag,dim01
> User: ADMIN
> Success: true
> Duration: 0.069
> Project: asiainfo_kylin
> Realization Names: [cube_1000100002_1]
> Cuboid Ids: [16]
> Total scan count: 6
> Result row count: 6
> Accept Partial: true
> Is Partial Result: false
> Hit Exception Cache: false
> Storage cache used: false
> Message: null
>
>
>
>
>
>
> nidongdong1987
--
Best regards,
Shaofeng Shi