[DRILL-1420] Hive query fails with uncaught Exception

• Type: Bug
• Status: Resolved
• Priority: Critical
• Resolution: Duplicate
• Affects Version/s: 0.5.0
• Fix Version/s: 0.5.0
• Component/s: Functions - Hive
• Assignee: Unassigned
• Reporter: Fabian Wilckens
• Created: 15/Sep/14 06:15
• Updated: 09/Dec/14 17:50
• Resolved: 15/Sep/14 07:51

Description

Hi Drill-Team,

everytime I try to query a hive table from drill-0.5.0 the following happens:

0: jdbc:drill:zk=hadoop2:5181,hadoop3:5181,ha> use hive;
----------------------+

ok

summary

----------------------+

true	Default schema changed to 'hive'

----------------------+
1 row selected (0,024 seconds)
0: jdbc:drill:zk=hadoop2:5181,hadoop3:5181,ha> show tables;
------------------------+

TABLE_SCHEMA

TABLE_NAME

------------------------+

hive.default

bondaten

------------------------+
1 row selected (0,09 seconds)
0: jdbc:drill:zk=hadoop2:5181,hadoop3:5181,ha> describe bondaten;
----------------------------------

COLUMN_NAME

DATA_TYPE

IS_NULLABLE

----------------------------------

id	VARCHAR	YES
type	INTEGER	YES

----------------------------------
2 rows selected (0,092 seconds)
0: jdbc:drill:zk=hadoop2:5181,hadoop3:5181,ha> select * from bondaten limit 10;
Query failed: Failure due to uncaught exception Instantiation of [simple type, class org.apache.drill.exec.store.hive.HiveTable$HivePartition] value failed: null (through reference chain: org.apache.drill.exec.store.hive.HivePartition["parameters"]) [5128c99b-0dfb-46ae-913d-7baf5542df3e]

Error: exception while executing query: Failure while trying to get next result batch. (state=,code=0)

Issue Links

• is duplicated by: DRILL-1398 Clone HivePartition only when InputSplit has non-null partition in InputSplit->Partition map