[
https://issues.apache.org/jira/browse/HIVE-20630?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
albertoramon updated HIVE-20630:
--------------------------------
Description:
In this code, the limit=1000 doesn't have expected result
{code:java}
INSERT OVERWRITE TABLE myNewTable partition (col1, col2)
select
. . .
from mySource
distribute by (floor(myColumn/2646 - + 1)/ 20 ) sort by merchant_fk
limit 1000;
select count (1) from ft_informes_trafico_stats_hdfs_arp;
>> 3001
{code}
was:
In this code, the limit=1000 doesn't have expected result
```sql
INSERT OVERWRITE TABLE myNewTable partition (col1, col2)
select
. . .
from mySource
distribute by (floor(myColumn/2646 - + 1)/ 20 ) sort by merchant_fk
limit 1000;
select count (1) from ft_informes_trafico_stats_hdfs_arp;
>> 3001
```
> Limit don't apply on Distribute clause
> --------------------------------------
>
> Key: HIVE-20630
> URL: https://issues.apache.org/jira/browse/HIVE-20630
> Project: Hive
> Issue Type: Bug
> Components: Query Planning
> Affects Versions: 2.3.1
> Reporter: albertoramon
> Priority: Minor
>
> In this code, the limit=1000 doesn't have expected result
>
> {code:java}
> INSERT OVERWRITE TABLE myNewTable partition (col1, col2)
> select
> . . .
> from mySource
> distribute by (floor(myColumn/2646 - + 1)/ 20 ) sort by merchant_fk
> limit 1000;
> select count (1) from ft_informes_trafico_stats_hdfs_arp;
> >> 3001
> {code}
--
This message was sent by Atlassian JIRA
(v7.6.3#76005)
