[CARBONDATA-2661] Query failed with group by column when MV Datamap created without group by column

• Type: Bug
• Status: Open
• Priority: Major
• Resolution: Unresolved
• Assignee: Bo Xu
• Reporter: Bo Xu
• Created: 27/Jun/18 02:43
• Updated: 09/Jul/18 13:51

Description

drop table if exists fact_table1;
CREATE TABLE fact_table1 (empno int, empname String, designation String, doj Timestamp,
workgroupcategory int, workgroupcategoryname String, deptno int, deptname String,
projectcode int, projectjoindate Timestamp, projectenddate Timestamp,attendance int,
utilization int,salary int)
STORED BY 'org.apache.carbondata.format';
  
  LOAD DATA local inpath 'hdfs://hacluster/user/hive/warehouse//data_mv.csv' INTO TABLE fact_table1 OPTIONS('DELIMITER'= ',', 'QUOTECHAR'= '"','timestampformat'='dd-MM-yyyy');

  
  LOAD DATA local inpath 'hdfs://hacluster/user/hive/warehouse//data_mv.csv' INTO TABLE fact_table1 OPTIONS('DELIMITER'= ',', 'QUOTECHAR'= '"','timestampformat'='dd-MM-yyyy');


0: jdbc:hive2://hadoop1:10000> create datamap mv1 using 'mv' as select sum(salary),count(empno) from fact_table1 group by empname;
+---------+--+
| Result  |
+---------+--+
+---------+--+
No rows selected (0.399 seconds)
0: jdbc:hive2://hadoop1:10000> rebuild datamap mv1;
+---------+--+
| Result  |
+---------+--+
+---------+--+
No rows selected (1.57 seconds)
0: jdbc:hive2://hadoop1:10000> create datamap mv2 using 'mv' as select sum(salary) from fact_table1 group by empname;
+---------+--+
| Result  |
+---------+--+
+---------+--+
No rows selected (0.321 seconds)
0: jdbc:hive2://hadoop1:10000> rebuild datamap mv2;
+---------+--+
| Result  |
+---------+--+
+---------+--+
No rows selected (1.241 seconds)
0: jdbc:hive2://hadoop1:10000> show datamap on table fact_table1;
+--------------+------------+--------------------------+--+
| DataMapName  | ClassName  |     Associated Table     |
+--------------+------------+--------------------------+--+
| datamap25    | mv         | default.datamap25_table  |
| mv1          | mv         | default.mv1_table        |
| mv2          | mv         | default.mv2_table        |
+--------------+------------+--------------------------+--+
3 rows selected (0.047 seconds)
0: jdbc:hive2://hadoop1:10000> select * from default.mv1_table;
+-------------+--------------+--+
| sum_salary  | count_empno  |
+-------------+--------------+--+
| 172332      | 18           |
| 162972      | 18           |
| 90720       | 18           |
| 202572      | 18           |
| 90720       | 18           |
| 128232      | 18           |
| 130410      | 18           |
| 202464      | 18           |
| 243846      | 18           |
| 238410      | 18           |
+-------------+--------------+--+
10 rows selected (0.314 seconds)
0: jdbc:hive2://hadoop1:10000> select count(*) from default.mv1_table;
+-----------+--+
| count(1)  |
+-----------+--+
| 10        |
+-----------+--+
1 row selected (0.139 seconds)
0: jdbc:hive2://hadoop1:10000> select * from default.mv2_table;
+-------------+--+
| sum_salary  |
+-------------+--+
| 172332      |
| 162972      |
| 90720       |
| 202464      |
| 243846      |
| 128232      |
| 130410      |
| 90720       |
| 202572      |
| 238410      |
+-------------+--+
10 rows selected (0.262 seconds)
0: jdbc:hive2://hadoop1:10000> explain select sum(salary) from fact_table1 group by empname;
+------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------+--+
|                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                            plan                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                            |
+------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------+--+
| == CarbonData Profiler ==
Table Scan on fact_table1
 - total blocklets: 2
 - filter: none
 - pruned by Main DataMap
    - skipped blocklets: 0
                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                            |
| == Physical Plan ==
*HashAggregate(keys=[empname#2499], functions=[sum(cast(salary#2511 as bigint))])
+- Exchange hashpartitioning(empname#2499, 200)
   +- *HashAggregate(keys=[empname#2499], functions=[partial_sum(cast(salary#2511 as bigint))])
      +- *BatchedScan CarbonDatasourceHadoopRelation [ Database name :default, Table name :fact_table1, Schema :Some(StructType(StructField(empno,IntegerType,true), StructField(empname,StringType,true), StructField(designation,StringType,true), StructField(doj,TimestampType,true), StructField(workgroupcategory,IntegerType,true), StructField(workgroupcategoryname,StringType,true), StructField(deptno,IntegerType,true), StructField(deptname,StringType,true), StructField(projectcode,IntegerType,true), StructField(projectjoindate,TimestampType,true), StructField(projectenddate,TimestampType,true), StructField(attendance,IntegerType,true), StructField(utilization,IntegerType,true), StructField(salary,IntegerType,true))) ] default.fact_table1[empname#2499,salary#2511]  |
+------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------+--+
2 rows selected (0.164 seconds)
0: jdbc:hive2://hadoop1:10000> explain select empname,sum(salary) from fact_table1 group by empname;
Error: org.apache.spark.sql.AnalysisException: cannot resolve '`fact_table1.empname`' given input columns: [sum_salary, count_empno]; line 6 pos 11;
'Aggregate ['fact_table1.empname], ['fact_table1.empname, 'sum('gen_subsumer_0.sum(salary)) AS sum(salary)#2802]
+- 'SubqueryAlias gen_subsumer_0
   +- 'Aggregate ['fact_table1.empname], [sum(sum_salary#2495L) AS sum(salary)#2800L, sum(count_empno#2496L) AS count(empno)#2801L]
      +- SubqueryAlias mv1_table
         +- Relation[sum_salary#2495L,count_empno#2496L] CarbonDatasourceHadoopRelation [ Database name :default, Table name :mv1_table, Schema :Some(StructType(StructField(sum_salary,LongType,true), StructField(count_empno,LongType,true))) ] (state=,code=0)
0: jdbc:hive2://hadoop1:10000> create datamap mv3 using 'mv' as select empname,sum(salary) from fact_table1 group by empname;
+---------+--+
| Result  |
+---------+--+
+---------+--+
No rows selected (0.318 seconds)
0: jdbc:hive2://hadoop1:10000> rebuild datamap mv3;
+---------+--+
| Result  |
+---------+--+
+---------+--+
No rows selected (1.437 seconds)
0: jdbc:hive2://hadoop1:10000> drop datamap mv3;
+---------+--+
| Result  |
+---------+--+
+---------+--+
No rows selected (0.941 seconds)
0: jdbc:hive2://hadoop1:10000> create datamap mv3 using 'mv' as select empname,sum(salary),count(empno) from fact_table1 group by empname;
+---------+--+
| Result  |
+---------+--+
+---------+--+
No rows selected (0.537 seconds)
0: jdbc:hive2://hadoop1:10000> create datamap mv4 using 'mv' as select empname,sum(salary) from fact_table1 group by empname;
+---------+--+
| Result  |
+---------+--+
+---------+--+
No rows selected (0.271 seconds)
0: jdbc:hive2://hadoop1:10000> rebuild datamap mv3;
+---------+--+
| Result  |
+---------+--+
+---------+--+
No rows selected (1.525 seconds)
0: jdbc:hive2://hadoop1:10000> rebuild datamap mv4;
+---------+--+
| Result  |
+---------+--+
+---------+--+
No rows selected (0.993 seconds)
0: jdbc:hive2://hadoop1:10000> explain select empname,sum(salary) from fact_table1 group by empname;
Error: org.apache.spark.sql.AnalysisException: cannot resolve '`fact_table1.empname`' given input columns: [sum_salary, count_empno]; line 6 pos 11;
'Aggregate ['fact_table1.empname], ['fact_table1.empname, 'sum('gen_subsumer_0.sum(salary)) AS sum(salary)#3113]
+- 'SubqueryAlias gen_subsumer_0
   +- 'Aggregate ['fact_table1.empname], [sum(sum_salary#2495L) AS sum(salary)#3111L, sum(count_empno#2496L) AS count(empno)#3112L]
      +- SubqueryAlias mv1_table
         +- Relation[sum_salary#2495L,count_empno#2496L] CarbonDatasourceHadoopRelation [ Database name :default, Table name :mv1_table, Schema :Some(StructType(StructField(sum_salary,LongType,true), StructField(count_empno,LongType,true))) ] (state=,code=0)

Issue Links

• is a child of: CARBONDATA-2628 Fix MV defects