[DRILL-2010] merge join returns wrong number of rows with large dataset

• Type: Bug
• Status: Closed
• Priority: Critical
• Resolution: Cannot Reproduce
• Affects Version/s: 0.8.0
• Fix Version/s: 1.15.0
• Component/s: Execution - Relational Operators
• Assignee: Boaz Ben-Zvi
• Reporter: Chun Chang
• Created: 14/Jan/15 00:48
• Updated: 20/Feb/19 02:57
• Resolved: 20/Feb/19 02:57

Description

#Mon Jan 12 18:19:31 EST 2015
git.commit.id.abbrev=5b012bf

When data set is big enough (like larger than one batch size), merge join will not returns the correct number of rows. Hash join returns the correct number of rows. Data can be downloaded from:

https://s3.amazonaws.com/apache-drill/files/complex100k.json.gz

With this dataset, the following query should return 10,000,000.

0: jdbc:drill:schema=dfs.drillTestDirComplexJ> alter session set `planner.enable_mergejoin` = true;
+------------+------------+
|     ok     |  summary   |
+------------+------------+
| true       | planner.enable_mergejoin updated. |
+------------+------------+
1 row selected (0.024 seconds)
0: jdbc:drill:schema=dfs.drillTestDirComplexJ> alter session set `planner.enable_hashjoin` = false;
+------------+------------+
|     ok     |  summary   |
+------------+------------+
| true       | planner.enable_hashjoin updated. |
+------------+------------+
1 row selected (0.024 seconds)
0: jdbc:drill:schema=dfs.drillTestDirComplexJ> select count(a.id) from `complex100k.json` a inner join `complex100k.json` b on a.gbyi=b.gbyi;
+------------+
|   EXPR$0   |
+------------+
| 9046760    |
+------------+
1 row selected (6.205 seconds)
0: jdbc:drill:schema=dfs.drillTestDirComplexJ> alter session set `planner.enable_mergejoin` = false;
+------------+------------+
|     ok     |  summary   |
+------------+------------+
| true       | planner.enable_mergejoin updated. |
+------------+------------+
1 row selected (0.026 seconds)
0: jdbc:drill:schema=dfs.drillTestDirComplexJ> alter session set `planner.enable_hashjoin` = true;
+------------+------------+
|     ok     |  summary   |
+------------+------------+
| true       | planner.enable_hashjoin updated. |
+------------+------------+
1 row selected (0.024 seconds)
0: jdbc:drill:schema=dfs.drillTestDirComplexJ> select count(a.id) from `complex100k.json` a inner join `complex100k.json` b on a.gbyi=b.gbyi;
+------------+
|   EXPR$0   |
+------------+
| 10000000   |
+------------+
1 row selected (4.453 seconds)

With smaller dataset, both merge and hash join returns the same correct number.

physical plan for merge join:

0: jdbc:drill:schema=dfs.drillTestDirComplexJ> explain plan for select count(a.id) from `complex100k.json` a inner join `complex100k.json` b on a.gbyi=b.gbyi;
+------------+------------+
|    text    |    json    |
+------------+------------+
| 00-00    Screen
00-01      StreamAgg(group=[{}], EXPR$0=[COUNT($0)])
00-02        Project(id=[$1])
00-03          MergeJoin(condition=[=($0, $2)], joinType=[inner])
00-05            SelectionVectorRemover
00-07              Sort(sort0=[$0], dir0=[ASC])
00-09                Scan(groupscan=[EasyGroupScan [selectionRoot=/drill/testdata/complex_type/json/complex100k.json, numFiles=1, columns=[`gbyi`, `id`], files=[maprfs:/drill/testdata/complex_type/json/complex100k.json]]])
00-04            Project(gbyi0=[$0])
00-06              SelectionVectorRemover
00-08                Sort(sort0=[$0], dir0=[ASC])
00-10                  Scan(groupscan=[EasyGroupScan [selectionRoot=/drill/testdata/complex_type/json/complex100k.json, numFiles=1, columns=[`gbyi`], files=[maprfs:/drill/testdata/complex_type/json/complex100k.json]]])