Spark Interpreter

Apache Spark is supported in Zeppelin with Spark Interpreter group, which consisted of 4 interpreters.

Name Class Description
%spark SparkInterpreter Creates SparkContext and provides scala environment
%pyspark PySparkInterpreter Provides python environment
%sql SparkSQLInterpreter Provides SQL environment
%dep DepInterpreter Dependency loader


Without any configuration, Spark interpreter works out of box in local mode. But if you want to connect to your Spark cluster, you'll need following two simple steps.

1. export SPARK_HOME

In conf/, export SPARK_HOME environment variable with your Spark installation path.

for example

export SPARK_HOME=/usr/lib/spark

You can optionally export HADOOP_CONF_DIR and SPARK_SUBMIT_OPTIONS

export HADOOP_CONF_DIR=/usr/lib/hadoop
export SPARK_SUBMIT_OPTIONS="--packages com.databricks:spark-csv_2.10:1.2.0"

2. set master in Interpreter menu.

After start Zeppelin, go to Interpreter menu and edit master property in your Spark interpreter setting. The value may vary depending on your Spark cluster deployment type.

for example,

  • local[*] in local mode
  • spark://master:7077 in standalone cluster
  • yarn-client in Yarn client mode
  • mesos://host:5050 in Mesos cluster

That's it. Zeppelin will work with any version of Spark and any deployment type without rebuild Zeppelin in this way. (Zeppelin 0.5.6-incubating release works up to Spark 1.6.1)

Note that without exporting SPARK_HOME, it's running in local mode with included version of Spark. The included version may vary depending on the build profile.

SparkContext, SQLContext, ZeppelinContext

SparkContext, SQLContext, ZeppelinContext are automatically created and exposed as variable names 'sc', 'sqlContext' and 'z', respectively, both in scala and python environments.

Note that scala / python environment shares the same SparkContext, SQLContext, ZeppelinContext instance.

Dependency Management

There are two ways to load external library in spark interpreter. First is using Zeppelin's %dep interpreter and second is loading Spark properties.

1. Dynamic Dependency Loading via %dep interpreter

When your code requires external library, instead of doing download/copy/restart Zeppelin, you can easily do following jobs using %dep interpreter.

  • Load libraries recursively from Maven repository
  • Load libraries from local filesystem
  • Add additional maven repository
  • Automatically add libraries to SparkCluster (You can turn off)

Dep interpreter leverages scala environment. So you can write any Scala code here. Note that %dep interpreter should be used before %spark, %pyspark, %sql.

Here's usages.

z.reset() // clean up previously added artifact and repository

// add maven repository

// add maven snapshot repository

// add credentials for private maven repository

// add artifact from filesystem

// add artifact from maven repository, with no dependency

// add artifact recursively

// add artifact recursively except comma separated GroupID:ArtifactId list
z.load("groupId:artifactId:version").exclude("groupId:artifactId,groupId:artifactId, ...")

// exclude with pattern

// local() skips adding artifact to spark clusters (skipping sc.addJar())

2. Loading Spark Properties

Once SPARK_HOME is set in conf/, Zeppelin uses spark-submit as spark interpreter runner. spark-submit supports two ways to load configurations. The first is command line options such as --master and Zeppelin can pass these options to spark-submit by exporting SPARK_SUBMIT_OPTIONS in conf/ Second is reading configuration options from SPARK_HOME/conf/spark-defaults.conf. Spark properites that user can set to distribute libraries are:

spark-defaults.conf SPARK_SUBMIT_OPTIONS Applicable Interpreter Description
spark.jars --jars %spark Comma-separated list of local jars to include on the driver and executor classpaths.
spark.jars.packages --packages %spark Comma-separated list of maven coordinates of jars to include on the driver and executor classpaths. Will search the local maven repo, then maven central and any additional remote repositories given by --repositories. The format for the coordinates should be groupId:artifactId:version.
spark.files --files %pyspark Comma-separated list of files to be placed in the working directory of each executor.

Note that adding jar to pyspark is only availabe via %dep interpreter at the moment

Here are few examples:


    export SPARK_SUBMIT_OPTIONS="--packages com.databricks:spark-csv_2.10:1.2.0 --jars /path/mylib1.jar,/path/mylib2.jar --files /path/,/path/,/path/mylib3.egg"
  • SPARK_HOME/conf/spark-defaults.conf

    spark.jars              /path/mylib1.jar,/path/mylib2.jar
    spark.jars.packages     com.databricks:spark-csv_2.10:1.2.0
    spark.files             /path/,/path/mylib2.egg,/path/


Zeppelin automatically injects ZeppelinContext as variable 'z' in your scala/python environment. ZeppelinContext provides some additional functions and utility.

Object exchange

ZeppelinContext extends map and it's shared between scala, python environment. So you can put some object from scala and read it from python, vise versa.

Put object from scala

val myObject = ...
z.put("objName", myObject)

Get object from python

myObject = z.get("objName")

Form creation

ZeppelinContext provides functions for creating forms. In scala and python environments, you can create forms programmatically.

/* Create text input form */

/* Create text input form with default value */
z.input("formName", "defaultValue")

/* Create select form */"formName", Seq(("option1", "option1DisplayName"),
                         ("option2", "option2DisplayName")))

/* Create select form with default value*/"formName", "option1", Seq(("option1", "option1DisplayName"),
                                    ("option2", "option2DisplayName")))

In sql environment, you can create form in simple template.

select * from ${table=defaultTableName} where text like '%${search}%'

To learn more about dynamic form, checkout Dynamic Form.