[HIVE-17259] Hive JDBC does not recognize UNIONTYPE columns

• Type: Bug
• Status: Closed
• Priority: Major
• Resolution: Fixed
• Fix Version/s: 3.0.0
• Component/s: Beeline, JDBC
• Labels: pull-request-available

Environment

Hive 1.2.1000.2.6.1.0-129
Beeline version 1.2.1000.2.6.1.0-129 by Apache Hive

• Assignee: Pierre Villard
• Reporter: Pierre Villard
• Created: 06/Aug/17 21:30
• Updated: 22/May/18 23:58
• Resolved: 07/Nov/17 18:35

Description

Hive JDBC does not recognize UNIONTYPE columns.

I've an external table backed by an avro schema containing a union type field.

"name" : "value",
"type" : [ "int", "string", "null" ]

When describing the table I've:

describe test_table;
+-------------------+---------------------------------------------------------------+----------+--+
|     col_name      |                           data_type                           | comment  |
+-------------------+---------------------------------------------------------------+----------+--+
| description       | string                                                        |          |
| name              | string                                                        |          |
| value             | uniontype<int,string,void>                                    |          |
+-------------------+---------------------------------------------------------------+----------+--+

When doing a select query over the data using the Hive CLI, it works:

hive> select value from test_table;
OK
{0:10}
{0:10}
{0:9}
{0:9}
...

But when using beeline, it fails:

0: jdbc:hive2://....> select * from test_table;
Error: Unrecognized column type: UNIONTYPE (state=,code=0)

By applying the patch provided with this JIRA, the command succeeds and return the expected output.

Issue Links

• links to: GitHub Pull Request #223