0

I am trying to analyze data stored in DSE 4.8.3 Cassandra using Tableau through SparkSQL.

Following is the description of column family I Created in Cassandra:

CREATE TABLE demokeyspace.temp (
    name text PRIMARY KEY,
    id int
)

The above table has only 1 row:

cqlsh> select * from demokeyspace.temp;

 name       | id
 ------------+----
 helloworld |  1

(1 rows)

One piece of info: I have replicated 'demokeyspace' keyspace on all the nodes of my cluster to make sure all the nodes can access it.

I have installed ODBC Administrator and Simba Spark ODBC Driver into the list of Drivers. I have enabled the thrift server on the node (spark5.xxx.xxx)

I am able to connect to the my Cassandra table from Tableau using SparkSQL.

My issue is when I drag the temp to the workspace and then click on 'Update Now' I am getting the following error:

enter image description here

When I click on the 'Show Query':

SELECT 1 AS `number_of_records`,
  `temp`.`id` AS `id`,
  `temp`.`name` AS `name`
FROM `demokeyspace`.`temp` `temp`
LIMIT 1000

Any idea why I am finding this error and/ or how I can fix this.

New Coder
  • 499
  • 4
  • 22
  • Can you share the version of the Spark driver you're using, along with a driver trace (see the install guide for the driver) of the problematic run? – KylePorter Feb 29 '16 at 18:37
  • I am using Spark version 1.4.2 – New Coder Mar 01 '16 at 22:38
  • I want to add 1 more thing. This error still persists. But I created another keyspace named 'tableaudemo' and created the same column family in that keyspace. It worked. Now I am not sure why this is running for the new keyspace but giving the same error for the 'demokeyspace' keyspace. – New Coder Mar 01 '16 at 22:41
  • Interesting. If you're able to provide the driver log I may be able to shed some light on this. – KylePorter Mar 03 '16 at 00:18

0 Answers0