[SPARK-20954] DESCRIBE showing 1 extra row of "| # col_name | data_type | comment |"

• Type: Bug
• Status: Resolved
• Priority: Major
• Resolution: Fixed
• Affects Version/s: 2.2.0
• Fix Version/s: 2.2.0
• Component/s: SQL
• Assignee: Dongjoon Hyun
• Reporter: Garros Chan
• Created: 01/Jun/17 18:46
• Updated: 15/Jun/17 00:39
• Resolved: 09/Jun/17 04:03

Description

I am trying to do DESCRIBE on a table but seeing 1 extra row being auto-added to the result. You can see there is this 1 extra row with "| # col_name | data_type | comment |" ; however, select and select count only shows 1 row.

I searched online a long time and do not find any useful information.
Is this a bug?

hdp106m2:/usr/hdp/2.5.0.2-3/spark2 # ./bin/beeline
Beeline version 1.2.1.spark2 by Apache Hive
[INFO] Unable to bind key for unsupported operation: backward-delete-word
[INFO] Unable to bind key for unsupported operation: backward-delete-word
[INFO] Unable to bind key for unsupported operation: down-history
[INFO] Unable to bind key for unsupported operation: up-history
[INFO] Unable to bind key for unsupported operation: up-history
[INFO] Unable to bind key for unsupported operation: down-history
[INFO] Unable to bind key for unsupported operation: up-history
[INFO] Unable to bind key for unsupported operation: down-history
[INFO] Unable to bind key for unsupported operation: up-history
[INFO] Unable to bind key for unsupported operation: down-history
[INFO] Unable to bind key for unsupported operation: up-history
[INFO] Unable to bind key for unsupported operation: down-history
beeline> !connect jdbc:hive2://localhost:10016
Connecting to jdbc:hive2://localhost:10016
Enter username for jdbc:hive2://localhost:10016: hive
Enter password for jdbc:hive2://localhost:10016: ****
17/06/01 14:13:04 INFO Utils: Supplied authorities: localhost:10016
17/06/01 14:13:04 INFO Utils: Resolved authority: localhost:10016
17/06/01 14:13:04 INFO HiveConnection: Will try to open client transport with JDBC Uri: jdbc:hive2://localhost:10016
Connected to: Spark SQL (version 2.2.1-SNAPSHOT)
Driver: Hive JDBC (version 1.2.1.spark2)
Transaction isolation: TRANSACTION_REPEATABLE_READ

0: jdbc:hive2://localhost:10016> describe garros.hivefloat;
-------------------------------+

col_name

data_type

comment

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

col_name	data_type	comment
c1	float	NULL

-------------------------------+
2 rows selected (0.396 seconds)

0: jdbc:hive2://localhost:10016> select * from garros.hivefloat;
---------------------+

c1

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

123.99800109863281

---------------------+
1 row selected (0.319 seconds)

0: jdbc:hive2://localhost:10016> select count from garros.hivefloat;
-----------+

count(1)

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

1

-----------+
1 row selected (0.783 seconds)

0: jdbc:hive2://localhost:10016> describe formatted garros.hiveint;
--------------------------------------------------------------------------------------------------+

col_name

data_type

comment

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

col_name	data_type	comment
c1	int	NULL
Detailed Table Information
Database	garros
Table	hiveint
Owner	root
Created	Thu Feb 09 17:40:36 EST 2017
Last Access	Wed Dec 31 19:00:00 EST 1969
Type	MANAGED
Provider	hive
Properties	[serialization.format=1]
Location	hdfs://HDP106/apps/hive/warehouse/garros.db/hiveint
Serde Library	org.apache.hadoop.hive.serde2.lazy.LazySimpleSerDe
InputFormat	org.apache.hadoop.mapred.TextInputFormat
OutputFormat	org.apache.hadoop.hive.ql.io.HiveIgnoreKeyTextOutputFormat
Partition Provider	Catalog

--------------------------------------------------------------------------------------------------+
17 rows selected (0.304 seconds)

Issue Links

• relates to: SPARK-20067 Unify and Clean Up Desc Commands Using Catalog Interface
• links to: [Github] Pull Request #18203 (dongjoon-hyun)
• links to: [Github] Pull Request #18245 (dongjoon-hyun)