[SPARK-11148] Unable to create views

• Type: Bug
• Status: Resolved
• Priority: Critical
• Resolution: Duplicate
• Affects Version/s: 1.5.1
• Fix Version/s: 2.0.0
• Component/s: SQL

Environment

Ubuntu 14.04
Spark-1.5.1-bin-hadoop2.6
(I don't have Hadoop or Hive installed)
Start spark-all.sh and thriftserver with mysql jar driver

• Assignee: Unassigned
• Reporter: Lunen
• Created: 16/Oct/15 15:54
• Updated: 04/May/16 03:24
• Resolved: 27/Apr/16 17:29

Description

I am unable to create views within spark SQL.
Creating tables without specifying the column names work. eg.

CREATE TABLE trade2
USING org.apache.spark.sql.jdbc
OPTIONS (
url "jdbc:mysql://192.168.30.191:3318/?user=root",
dbtable "database.trade",
driver "com.mysql.jdbc.Driver"
);

Ceating tables with datatypes gives an error:

CREATE TABLE trade2(
COL1 timestamp,
COL2 STRING,
COL3 STRING)
USING org.apache.spark.sql.jdbc
OPTIONS (
url "jdbc:mysql://192.168.30.191:3318/?user=root",
dbtable "database.trade",
driver "com.mysql.jdbc.Driver"
);
Error: org.apache.spark.sql.AnalysisException: org.apache.spark.sql.execution.datasources.jdbc.DefaultSource does not allow user-specified schemas.; SQLState: null ErrorCode: 0

Trying to create a VIEW from the table that was created.(The select statement below returns data)
CREATE VIEW viewtrade as Select Col1 from trade2;

Error: org.apache.spark.sql.execution.QueryExecutionException: FAILED: SemanticException [Error 10004]: Line 1:30 Invalid table alias or column reference 'Col1': (possible column names are: col)
SQLState: null
ErrorCode: 0

Issue Links

• duplicates: SPARK-11012 Canonicalize view definitions
• is related to: SPARK-11012 Canonicalize view definitions