3

i have zeppelin running on my system.. details are

ubuntu 14.04 64bit

Apache Zeppelin 0.6.0

java version "1.7.0_111"

Apache Maven 3.3.9

-Pspark-2.0 -Dspark.version=1.5.2

i have seen many interpreters with sql but i didnot get which interpreter to go with. all like ignite,jdbc,live,psql,spark interpreters are not providing proper url, username, password, database to connect my MS sql server.

i read some posts and assume that there is some interpreter which connects zeppelin-sql server. Apache zeppelin tutorial, error "sql interpreter not found" this also didnt work. could someone please guide me to select the interpreter if exist, if not how to add this interpreter?

thanks in advance

Community
  • 1
  • 1
Priyanka D L
  • 171
  • 1
  • 8
  • I'm not sure if there is such interpreter. The SQL interpreter you mentioned isn't Spark SQL interpreter and not MS SQL server's interpreter. – eliasah Aug 25 '16 at 20:07

2 Answers2

4

  • download JDBC driver for sql server provided by MS

  • exact sqljdbc42.jar and save it to local, e.g. /zeppelin/3rd-lib/sqljdbc42.jar

  • create new interpreter with MS SQLServer Driverenter image description here

  • use it in the notebook!

Xia Lingxue
  • 91
  • 1
  • 3
  • Wow!! This was ***so hugely helpful***. This answer was concise, explicit, and had every detail I needed to perform something that I thought would be very complex!! Kudos to @Xia!! – Mike Williamson Feb 02 '18 at 22:37
  • Have you tried this using windows authentication? Setting the auth type to be kerboros? – Michael Black May 03 '19 at 19:52
0

Apache Zeppelin doesn't come with a specific interpreter for SQL Server. By default Apache Zeppelin is shipped with a JDBC interpreter which means that you can use that to connect to SQL Server, provided that you have downloaded and installed the Microsoft SQL Server JDBC Driver.

To overcome the limitations of the generic JDBC interpreter I've created a specific Zeppelin Interpreter for SQL Server. You can find more details here:

https://medium.com/@mauridb/apache-zeppelin-sql-server-sql-azure-and-sql-azure-dw-1706697bb54#.wuk7dte00

You can use it to connect to SQL Server, Azure SQL and Azure DW

mauridb
  • 1,467
  • 9
  • 12
  • Hi @Davide, it looks like you did some great work on that link, but why would I need it? What does your interpreter provide **over** JDBC? Is it faster?? Does it allow for more specific MS Server commands? And the docker image you have, it looks like it's just to **host an MS SQL server**. But the OP, and I suspect most folks visiting here, want to *connect* to a server that already exists. – Mike Williamson Feb 02 '18 at 22:42
  • 1
    Hi @MikeWilliamson. Two things here: JDBC connection is not all that simple for someone not used to setting it up, especially under linux. You can easily send *hours* trying to have it running if you are not already experienced in the field. Second point my idea is to add SQL Server specific feature supports. For example to easily get more insight in query execution, especially from a performance perspective. Unfortunately I had to slow down a lot development lately – mauridb Mar 09 '18 at 23:44