Aman Sinha created DRILL-886:
--------------------------------
Summary: Wrong results for a query with Right Outer Join on the
second (and subsequent) executions
Key: DRILL-886
URL: https://issues.apache.org/jira/browse/DRILL-886
Project: Apache Drill
Issue Type: Bug
Reporter: Aman Sinha
The following query with a right outer join produces correct results on the
first execution in a session but wrong results on the second and subsequent
executions. A potential cause for the problem can be seen from the two
Explain plans: the scan of the nation table shows a difference in the columns
being projected.
0: jdbc:drill:zk=local> select n.n_regionkey, r.r_regionkey from
cp.`tpch/region.parquet` r right join cp.`tpch/nation.parquet` n on
n.n_regionkey = r.r_regionkey;
+-------------+-------------+
| n_regionkey | r_regionkey |
+-------------+-------------+
| 0 | 0 |
| 0 | 0 |
| 0 | 0 |
| 0 | 0 |
| 0 | 0 |
| 1 | 1 |
| 1 | 1 |
| 1 | 1 |
| 1 | 1 |
| 1 | 1 |
| 2 | 2 |
| 2 | 2 |
| 2 | 2 |
| 2 | 2 |
| 2 | 2 |
| 3 | 3 |
| 3 | 3 |
| 3 | 3 |
| 3 | 3 |
| 3 | 3 |
| 4 | 4 |
| 4 | 4 |
| 4 | 4 |
| 4 | 4 |
| 4 | 4 |
+-------------+-------------+
25 rows selected (2.207 seconds)
0: jdbc:drill:zk=local> select n.n_regionkey, r.r_regionkey from
cp.`tpch/region.parquet` r right join cp.`tpch/nation.parquet` n on
n.n_regionkey = r.r_regionkey;
+-------------+-------------+
| n_regionkey | r_regionkey |
+-------------+-------------+
| 0 | null |
| 1 | null |
| 1 | null |
| 1 | null |
| 4 | null |
| 0 | null |
| 3 | null |
| 3 | null |
| 2 | null |
| 2 | null |
| 4 | null |
| 4 | null |
| 2 | null |
| 4 | null |
| 0 | null |
| 0 | null |
| 0 | null |
| 1 | null |
| 2 | null |
| 3 | null |
| 4 | null |
| 2 | null |
| 3 | null |
| 3 | null |
| 1 | null |
+-------------+-------------+
25 rows selected (0.514 seconds)
EXPLAIN plan for the good run:
| 00-00 Screen
00-01 Project(n_regionkey=[$0], r_regionkey=[$1])
00-02 Project(n_regionkey=[$3], r_regionkey=[$1])
00-03 HashJoin(condition=[=($3, $1)], joinType=[right])
00-05 Scan(groupscan=[ParquetGroupScan [entries=[ReadEntryWithPath
[path=/tpch/region.parquet]], selectionRoot=/tpch/region.parquet,
columns=[SchemaPath [`r_regionkey`]]]])
00-04 Project(*0=[$0], n_regionkey=[$1])
00-06 BroadcastExchange
01-01 Scan(groupscan=[ParquetGroupScan
[entries=[ReadEntryWithPath [path=/tpch/nation.parquet]],
selectionRoot=/tpch/nation.parquet, columns=[SchemaPath [`n_regionkey`]]]])
Explain plan for the bad run:
| 00-00 Screen
00-01 Project(n_regionkey=[$0], r_regionkey=[$1])
00-02 Project(n_regionkey=[$3], r_regionkey=[$1])
00-03 HashJoin(condition=[=($2, $1)], joinType=[right])
00-05 Scan(groupscan=[ParquetGroupScan [entries=[ReadEntryWithPath
[path=/tpch/region.parquet]], selectionRoot=/tpch/region.parquet,
columns=[SchemaPath [`r_regionkey`]]]])
00-04 Project(*0=[$0], n_regionkey=[$1])
00-06 BroadcastExchange
01-01 Scan(groupscan=[ParquetGroupScan
[entries=[ReadEntryWithPath [path=/tpch/nation.parquet]],
selectionRoot=/tpch/nation.parquet, columns=null]])
--
This message was sent by Atlassian JIRA
(v6.2#6252)
