Use Apache Zeppelin notebooks with Apache Spark cluster on Azure HDInsight

HDInsight Spark clusters include Apache Zeppelin notebooks that you can use to run Apache Spark jobs. In this article, you learn how to use the Zeppelin notebook on an HDInsight cluster.


Launch an Apache Zeppelin notebook

  1. From the Spark cluster Overview, select Zeppelin notebook from Cluster dashboards. Enter the admin credentials for the cluster.


    You may also reach the Zeppelin Notebook for your cluster by opening the following URL in your browser. Replace CLUSTERNAME with the name of your cluster:

  2. Create a new notebook. From the header pane, navigate to Notebook > Create new note.

    Create a new Zeppelin notebook

    Enter a name for the notebook, then select Create Note.

  3. Ensure the notebook header shows a connected status. It's denoted by a green dot in the top-right corner.

    Zeppelin notebook status

  4. Load sample data into a temporary table. When you create a Spark cluster in HDInsight, the sample data file, hvac.csv, is copied to the associated storage account under \HdiSamples\SensorSampleData\hvac.

    In the empty paragraph that is created by default in the new notebook, paste the following snippet.

    //The above magic instructs Zeppelin to use the Livy Scala interpreter
    // Create an RDD using the default Spark context, sc
    val hvacText = sc.textFile("wasbs:///HdiSamples/HdiSamples/SensorSampleData/hvac/HVAC.csv")
    // Define a schema
    case class Hvac(date: String, time: String, targettemp: Integer, actualtemp: Integer, buildingID: String)
    // Map the values in the .csv file to the schema
    val hvac = => s.split(",")).filter(s => s(0) != "Date").map(
        s => Hvac(s(0),
    // Register as a temporary table called "hvac"

    Press SHIFT + ENTER or select the Play button for the paragraph to run the snippet. The status on the right-corner of the paragraph should progress from READY, PENDING, RUNNING to FINISHED. The output shows up at the bottom of the same paragraph. The screenshot looks like the following:

    Create a temporary table from raw data

    You can also provide a title to each paragraph. From the right-hand corner of the paragraph, select the Settings icon (sprocket), and then select Show title.


    %spark2 interpreter is not supported in Zeppelin notebooks across all HDInsight versions, and %sh interpreter will not be supported from HDInsight 4.0 onwards.

  5. You can now run Spark SQL statements on the hvac table. Paste the following query in a new paragraph. The query retrieves the building ID and the difference between the target and actual temperatures for each building on a given date. Press SHIFT + ENTER.

    select buildingID, (targettemp - actualtemp) as temp_diff, date from hvac where date = "6/1/13"

    The %sql statement at the beginning tells the notebook to use the Livy Scala interpreter.

  6. Select the Bar Chart icon to change the display. settings, which appears after you have selected Bar Chart, allows you to choose Keys, and Values. The following screenshot shows the output.

    Run a Spark SQL statement using the notebook1

  7. You can also run Spark SQL statements using variables in the query. The next snippet shows how to define a variable, Temp, in the query with the possible values you want to query with. When you first run the query, a drop-down is automatically populated with the values you specified for the variable.

    select buildingID, date, targettemp, (targettemp - actualtemp) as temp_diff from hvac where targettemp > "${Temp = 65,65|75|85}"

    Paste this snippet in a new paragraph and press SHIFT + ENTER. Then select 65 from the Temp drop-down list.

  8. Select the Bar Chart icon to change the display. Then select settings and make the following changes:

    • Groups: Add targettemp.

    • Values: 1. Remove date. 2. Add temp_diff. 3. Change the aggregator from SUM to AVG.

      The following screenshot shows the output.

      Run a Spark SQL statement using the notebook2

How do I use external packages with the notebook?

You can configure the Zeppelin notebook in Apache Spark cluster on HDInsight to use external, community-contributed packages that aren't included out-of-the-box in the cluster. You can search the Maven repository for the complete list of packages that are available. You can also get a list of available packages from other sources. For example, a complete list of community-contributed packages is available at Spark Packages.

In this article, you'll see how to use the spark-csv package with the Jupyter notebook.

  1. Open interpreter settings. From the top-right corner, select the logged in user name, then select Interpreter.

    Launch interpreter

  2. Scroll to livy2, then select edit.

    Change interpreter settings1

  3. Navigate to key livy.spark.jars.packages, and set its value in the format group:id:version. So, if you want to use the spark-csv package, you must set the value of the key to com.databricks:spark-csv_2.10:1.4.0.

    Change interpreter settings2

    Select Save and then OK to restart the Livy interpreter.

  4. If you want to understand how to arrive at the value of the key entered above, here's how.

    a. Locate the package in the Maven Repository. For this article, we used spark-csv.

    b. From the repository, gather the values for GroupId, ArtifactId, and Version.

    Use external packages with Jupyter notebook

    c. Concatenate the three values, separated by a colon (:).


Where are the Zeppelin notebooks saved?

The Zeppelin notebooks are saved to the cluster headnodes. So, if you delete the cluster, the notebooks will be deleted as well. If you want to preserve your notebooks for later use on other clusters, you must export them after you have finished running the jobs. To export a notebook, select the Export icon as shown in the image below.

Download notebook

This saves the notebook as a JSON file in your download location.

Livy session management

When you run the first code paragraph in your Zeppelin notebook, a new Livy session is created in your HDInsight Spark cluster. This session is shared across all Zeppelin notebooks that you subsequently create. If for some reason the Livy session is killed (cluster reboot, etc.), you will not be able to run jobs from the Zeppelin notebook.

In such a case, you must perform the following steps before you can start running jobs from a Zeppelin notebook.

  1. Restart the Livy interpreter from the Zeppelin notebook. To do so, open interpreter settings by selecting the logged in user name from the top-right corner, then select Interpreter.

    Launch interpreter

  2. Scroll to livy2, then select restart.

    Restart the Livy interpreter

  3. Run a code cell from an existing Zeppelin notebook. This creates a new Livy session in the HDInsight cluster.

See also


Create and run applications

Tools and extensions

Manage resources