[DRILL-5947] COUNT(<column-name>) on empty JSON results in schema change error

• Type: Bug
• Status: Open
• Priority: Major
• Resolution: Unresolved
• Affects Version/s: 1.11.0
• Component/s: Execution - Flow
• Assignee: Paul Rogers
• Reporter: Khurram Faraaz
• Created: 09/Nov/17 05:21
• Updated: 09/Nov/17 07:45

Description

Count over a field, from a non-empty and another empty JSON file results in schema change error, "Streaming aggregate does not support schema changes"

OpenJDK 64-Bit Server VM warning: ignoring option MaxPermSize=512M; support was removed in 8.0
apache drill 1.11.0-mapr 
"drill baby drill"
0: jdbc:drill:schema=dfs.tmp> select count(id) from `diff_types`;
Error: UNSUPPORTED_OPERATION ERROR: Streaming aggregate does not support schema changes
Prior schema : 
BatchSchema [fields=[id(BIGINT:OPTIONAL)], selectionVector=NONE]
New schema : 
BatchSchema [fields=[id(INT:OPTIONAL)], selectionVector=NONE]

Fragment 0:0

[Error Id: c1c77a0e-c7cf-494e-b8d2-ef2cc7e816e0 on centos-01.qa.lab:31010] (state=,code=0)
0: jdbc:drill:schema=dfs.tmp> select id from `dif`Display all 872 possibilities? (y or n)
0: jdbc:drill:schema=dfs.tmp> select id from `diff_types/m1.json`;
+-------+
|  id   |
+-------+
| 1234  |
+-------+
1 row selected (0.157 seconds)
0: jdbc:drill:schema=dfs.tmp> select id from `diff_types/m2.json`;
+-------+
|  id   |
+-------+
| null  |
+-------+
1 row selected (0.169 seconds)
0: jdbc:drill:schema=dfs.tmp> select * from `diff_types`;
+-------+
|  id   |
+-------+
| 1234  |
| null  |
+-------+
2 rows selected (0.209 seconds)
0: jdbc:drill:schema=dfs.tmp> select count(*) from `diff_types`;
+---------+
| EXPR$0  |
+---------+
| 2       |
+---------+
1 row selected (0.208 seconds)