hive-issues mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Vineet Garg (JIRA)" <j...@apache.org>
Subject [jira] [Commented] (HIVE-14522) CBO: Calcite Operator To Hive Operator(Calcite Return Path): Fix test failure for auto_join_filters
Date Fri, 12 Aug 2016 01:33:20 GMT

    [ https://issues.apache.org/jira/browse/HIVE-14522?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15418231#comment-15418231
] 

Vineet Garg commented on HIVE-14522:
------------------------------------

This isn't actually a wrong result this is correct result. This test is for HIVE-1534 where
outer join semantics were fixed and a hive configuration parameter was added to maintain backward
compatibility (with wrong result). This test is testing that backward compatibility. It seems
converting from calcite tree to hive operator tree this flag is being ignored. I am not sure
if it is worth supporting this backward compatibility in return path since it's been almost
6 years now.

> CBO: Calcite Operator To Hive Operator(Calcite Return Path): Fix test failure for auto_join_filters
> ---------------------------------------------------------------------------------------------------
>
>                 Key: HIVE-14522
>                 URL: https://issues.apache.org/jira/browse/HIVE-14522
>             Project: Hive
>          Issue Type: Sub-task
>          Components: CBO
>            Reporter: Vineet Garg
>            Assignee: Vineet Garg
>
> {code}
> CREATE TABLE smb_input1(key int, value int) CLUSTERED BY (key) SORTED BY (key) INTO 2
BUCKETS; 
> CREATE TABLE smb_input2(key int, value int) CLUSTERED BY (value) SORTED BY (value) INTO
2 BUCKETS; 
> LOAD DATA LOCAL INPATH '../../data/files/in1.txt' into table smb_input1;
> LOAD DATA LOCAL INPATH '../../data/files/in2.txt' into table smb_input1;
> LOAD DATA LOCAL INPATH '../../data/files/in1.txt' into table smb_input2;
> LOAD DATA LOCAL INPATH '../../data/files/in2.txt' into table smb_input2;
> SET hive.optimize.bucketmapjoin = true;
> SET hive.optimize.bucketmapjoin.sortedmerge = true;
> SET hive.input.format = org.apache.hadoop.hive.ql.io.BucketizedHiveInputFormat;
> SET hive.outerjoin.supports.filters = false;
> {code}
> {code} SELECT sum(hash(a.key,a.value,b.key,b.value)) FROM myinput1 a LEFT OUTER JOIN
myinput1 b on a.key > 40 AND a.value > 50 AND a.key = a.value AND b.key > 40 AND
b.value > 50 AND b.key = b.value; {code}
> {code} Expected result: 3078400 Actual result: 4937935 {code}



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

Mime
View raw message