I'm not sure you could use that syntax with the Hive metastore. It would be
LOAD DATA LOCAL PATH like here:
http://spark.apache.org/docs/latest/sql-programming-guide.html#hive-tables
_____________________________
From: moon soo Lee <m...@apache.org>
Sent: Saturday, November 7, 2015 10:12 PM
Subject: Re: sql interpreter
To: <users@zeppelin.incubator.apache.org>
Hi,
Give sql syntax looks like only works with "HiveContext".
Zeppelin tries to create "HiveContext" if possible, but when it couldn't, it
fall back to "SQLContext", which may not support such syntax.
So could you check your logs to see any related exception with
initializing HiveContext?
Best, moon
On Sun, Nov 1, 2015 at 4:25 PM Tomas Hudik <
xhu...@gmail.com> wrote:
Hi there,
I have a problem with sql interpreter:
%sql
CREATE TABLE bla
USING com.databricks.spark.csv
OPTIONS (path "bank-full.csv", header "true", delimiter ';',inferSchema =
"true")
(code taken from https://github.com/databricks/spark-csv ).
If I run it i got:
org.apache.spark.sql.AnalysisException: cannot recognize input near
'bla' 'USING' 'com' in table name; line 2 pos 0 at
org.apache.spark.sql.hive.HiveQl$.createPlan(HiveQl.scala:297)
Any idea why? Zeppelin's sql interperter cannot create table?
thanks, Tomas
