How To Use Zeppelin With SAP HANA

Apache Zeppelin is an open source tool that allows interactive data analytics from many data sources like databases, hive, spark, python, hdfs, HANA and more.

It allows for:

Data Ingestion
Data Discovery
Data Analytics
Data Visualization & Collaboration

httpsaws.amazon.com/blogs/aws/amazon-emr-update-apache-spark-1-5-2-ganglia-presto-zeppelin-and-oozie/

SAP HANA is an in memory data platform for storing and analyzing data. At its core it is a columner database with the ability to also become a row database.

HANA overview

Setup Zeppelin

Install Apache Zeppelin from https://zeppelin.apache.org/download.html with all interpreters. Once Zeppelin is installed go to the bin directory and run ./zeppelin-daemon.sh start You can view the Zeppelin interface at http://localhost:8080/

zeppelin start screen

Click on the menu to go to interpreter setup.

Configuration Menu

Choose the Create button.

Create new interpreter

Pick the JDBC interpreter group in the pull down.

Pick JDBC Interpreter

Fill in these HANA properties with your specific server name and port. Give it the name of hana.

HANA interpreter property values

The important parts are:

default.driver: com.sap.db.jdbc.Driver

default.url: jdbc:sap://[server]:[port]

Save the interpreter with the save button.

Install JDBC Driver

Find the jar file called ngdbc.jar and place it in zeppelin-0.x.X-bin-all/interpreter/jdbc.

The ngdbc.jar may be imbedded in another jar file. You will need to install the SAP HANA Studio or the sql client drivers. In order for me to find the ngdbc.jar I had to unzip the com.sap.ndb.studio.jdbc_2.3.6.jar file by changing the .jar to .zip. When it is unzipped I found the ngdbc.jar in the lib directory.

The com.sap.ndb.studio.jdbc_2.3.6.jar was found in a directory called studio/core_repository/plugins

Make a New Note

Go to the menu and create a new note.

Create New Note

Pick hana as your default interpreter and give your note a name.

Create New Note

This will create a workspace where you can create your tables and graphs from data in HANA and also annotate your report with text. It is also possible to pull in data from other database systems you have already configured.

Create New Note

Now enter your query in the text area called the paragraph.

type in query

Click the arrow to run the query and then you can select what kind of graph to display.

Type in query and pick a visualization

Now you can add additional paragraphs with markdown text in them to describe your information.

So we have successfully configured Apache Zeppelin to pull data from HANA using a jdbc driver.

How to Use HANA from Spark

You may also want to connect to HANA directly from Spark using Scala, Python or R code. The best way to implement this is to put the reference to the jdbc jar called ngdbc.jar in the Spark interpreter. Go to the Zeppelin settings menu.