[HIVE-22227] Tez bucket pruning produces wrong result with shared work optimization

• Type: Bug
• Status: Closed
• Priority: Major
• Resolution: Fixed
• Affects Version/s: 4.0.0
• Fix Version/s: 4.0.0-alpha-1
• Component/s: Query Planning
• Target Version/s: 4.0.0
• Assignee: Vineet Garg Assign to me
• Reporter: Vineet Garg
• Created: 20/Sep/19 22:09
• Updated: 27/Feb/24 22:23
• Resolved: 25/Sep/19 01:15

Description

Reproducer

set hive.tez.bucket.pruning=true;
set hive.optimize.shared.work=true;

CREATE TABLE srcbucket_mapjoin_n16(key int, value string) partitioned by (ds string) CLUSTERED BY (key) INTO 2 BUCKETS STORED AS TEXTFILE;
CREATE TABLE tab_part_n10 (key int, value string) PARTITIONED BY(ds STRING) CLUSTERED BY (key) SORTED BY (key) INTO 4 BUCKETS STORED AS ORCFILE;
CREATE TABLE srcbucket_mapjoin_part_n17 (key int, value string) partitioned by (ds string) CLUSTERED BY (key) INTO 4 BUCKETS STORED AS TEXTFILE;

load data local inpath '$HIVE_SRC/data/files/bmj/000000_0' INTO TABLE srcbucket_mapjoin_n16 partition(ds='2008-04-08');
load data local inpath '.$HIVE_SRC/data/files/bmj1/000001_0' INTO TABLE srcbucket_mapjoin_n16 partition(ds='2008-04-08');

load data local inpath '$HIVE_SRC/data/files/bmj/000000_0' INTO TABLE srcbucket_mapjoin_part_n17 partition(ds='2008-04-08');
load data local inpath '$HIVE_SRC/data/files/bmj/000001_0' INTO TABLE srcbucket_mapjoin_part_n17 partition(ds='2008-04-08');
load data local inpath '$HIVE_SRC/data/files/bmj/000002_0' INTO TABLE srcbucket_mapjoin_part_n17 partition(ds='2008-04-08');


set hive.optimize.bucketingsorting=false;
insert overwrite table tab_part_n10 partition (ds='2008-04-08')
select key,value from srcbucket_mapjoin_part_n17;

CREATE TABLE tab_n9(key int, value string) PARTITIONED BY(ds STRING) CLUSTERED BY (key) SORTED BY (key) INTO 2 BUCKETS STORED AS ORCFILE;
insert overwrite table tab_n9 partition (ds='2008-04-08')
select key,value from srcbucket_mapjoin_n16;

select * from
(select * from tab_n9 where tab_n9.key = 0)a
join
(select * from tab_part_n10 where tab_part_n10.key = 98)b full outer join tab_part_n10 c on a.key = b.key and b.key = c.key
order by 1,2,3,4,5,6,7,8,9;

Issue Links

• relates to: HIVE-22136 Turn on tez.bucket.pruning