Babulal created CARBONDATA-2550:
-----------------------------------
Summary: [MV] Limit is ignored when data fetched from MV, Query
rewrite is Wrong
Key: CARBONDATA-2550
URL: https://issues.apache.org/jira/browse/CARBONDATA-2550
Project: CarbonData
Issue Type: Bug
Reporter: Babulal
0: jdbc:hive2://10.18.222.231:23040> create table mvtable1(name string,age
int,salary int) stored by 'carbondata';
+---------+--+
| Result |
+---------+--+
+---------+--+
No rows selected (0.279 seconds)
0: jdbc:hive2://10.18.222.231:23040> insert into mvtable1 select 'n1',12,12;
+---------+--+
| Result |
+---------+--+
+---------+--+
No rows selected (11.973 seconds)
0: jdbc:hive2://10.18.222.231:23040> insert into mvtable1 select 'n1',12,12;
+---------+--+
| Result |
+---------+--+
+---------+--+
No rows selected (9.92 seconds)
0: jdbc:hive2://10.18.222.231:23040> insert into mvtable1 select 'n3',12,12;
+---------+--+
| Result |
+---------+--+
+---------+--+
No rows selected (9.883 seconds)
0: jdbc:hive2://10.18.222.231:23040> insert into mvtable1 select 'n4',12,12;
+---------+--+
| Result |
+---------+--+
+---------+--+
No rows selected (10.488 seconds)
0: jdbc:hive2://10.18.222.231:23040> select name,sum(salary) from mvtable1
group by name;
+-------+--------------+--+
| name | sum(salary) |
+-------+--------------+--+
| n3 | 12 |
| n1 | 24 |
| n4 | 12 |
+-------+--------------+–+
0: jdbc:hive2://10.18.222.231:23040> select name,sum(salary) from mvtable1
group by name limit 2;
+-------+--------------+--+
| name | sum(salary) |
+-------+--------------+--+
| n3 | 12 |
| n1 | 24 |
+-------+--------------+--+
2 rows selected (4.175 seconds)
0: jdbc:hive2://10.18.222.231:23040> create datamap map1 using 'mv' as select
name,sum(salary) from mvtable1 group by name;
+---------+--+
| Result |
+---------+--+
+---------+--+
No rows selected (0.396 seconds)
0: jdbc:hive2://10.18.222.231:23040> rebuild datamap map1;
+---------+--+
| Result |
+---------+--+
+---------+--+
No rows selected (13.246 seconds)
*0: jdbc:hive2://10.18.222.231:23040> select name,sum(salary) from mvtable1
group by name limit 2;*
+----------------+-------------+--+
| mvtable1_name | sum_salary |
+----------------+-------------+--+
| n3 | 12 |
| n1 | 24 |
| n4 | 12 |
+----------------+-------------+--+
3 rows selected (2.453 seconds)
*0: jdbc:hive2://10.18.222.231:23040> select name,sum(salary) from mvtable1
group by name limit 1;*
+----------------+-------------+--+
| mvtable1_name | sum_salary |
+----------------+-------------+--+
| n3 | 12 |
| n1 | 24 |
| n4 | 12 |
+----------------+-------------+--+
3 rows selected (0.347 seconds)
0: jdbc:hive2://10.18.222.231:23040>
Even limit is given MV returns all the records from MV table.
Cause:-
When Rewriting MV query ,limit is ignored,
0: jdbc:hive2://10.18.222.231:23040> explain select name,sum(salary) from
mvtable1 *group by name limit 2;*
| plan |
| == CarbonData Profiler ==
Table Scan on map1_table
- total blocklets: 2
- filter: none
- pruned by Main DataMap
- skipped blocklets: 0
|
| == Physical Plan ==
*BatchedScan CarbonDatasourceHadoopRelation [ Database name :default, Table
name :map1_table, Schema
:Some(StructType(StructField(mvtable1_name,StringType,true),
StructField(sum_salary,LongType,true))) ]
default.map1_table[mvtable1_name#4438,sum_salary#4614L] |
+------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------+--+
2 rows selected (0.36
--
This message was sent by Atlassian JIRA
(v7.6.3#76005)
