hive-issues mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Vikash Pareek (JIRA)" <j...@apache.org>
Subject [jira] [Comment Edited] (HIVE-15272) "LEFT OUTER JOIN" Is not populating correct records with Hive On Spark
Date Thu, 24 Nov 2016 15:11:00 GMT

    [ https://issues.apache.org/jira/browse/HIVE-15272?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15693243#comment-15693243
] 

Vikash Pareek edited comment on HIVE-15272 at 11/24/16 3:10 PM:
----------------------------------------------------------------

I am just calculating count of the records, result (count) does not dependent on ordering.
Result should be same for each execution as in case of MR.

my_table1 (left) is having ~30 million records 
my_table2 (right) is having ~85 million records 



was (Author: vpareek):
I am just calculating count of the records, result (count) does not dependent on ordering.
Result should be same for each execution as in case of MR.

I have around 30 million data in my_table1 (left) and 85 million data in my_table2 (right).


> "LEFT OUTER JOIN" Is not populating correct records with Hive On Spark
> ----------------------------------------------------------------------
>
>                 Key: HIVE-15272
>                 URL: https://issues.apache.org/jira/browse/HIVE-15272
>             Project: Hive
>          Issue Type: Bug
>          Components: Hive, Spark
>    Affects Versions: 1.1.0
>         Environment: Hive 1.1.0, CentOS, Cloudera 5.7.4
>            Reporter: Vikash Pareek
>
> I ran following Hive query multiple times with execution engine as Hive on Spark and
Hive on MapReduce.
> {code}
> SELECT COUNT(DISTINCT t1.region, t1.amount)
> FROM my_db.my_table1 t1
> LEFT OUTER
> JOIN my-db.my_table2 t2 ON (t1.id = t2.id
>                             AND t1.name = t2.name)
> {code}
> With Hive on Spark: Result (count) were different of every execution.
> With Hive on MapReduce: Result (count) were same of every execution.
> Seems like Hive on Spark behaving differently in each execution and does not populating
correct result.



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

Mime
View raw message