Install Apache Spark on Windows | Hostvento Knowledge Base

Before you start

Spark needs Java. Open Command Prompt and check that it is installed:

java --version

If the command is not found, install a JDK first.

Step 1: Download Spark

Go to the Apache Spark download page and pick a release and a package type. Download the .tgz file.

Step 2: Extract Spark

Extract the archive to a short path such as C:\spark. A short path with no spaces avoids many errors.

Step 3: Set the environment variables

Search for Edit the system environment variables in the Start menu and open it.
Click Environment Variables.
Under System variables, click New. Set the name to SPARK_HOME and the value to your Spark folder, for example C:\spark.
Select Path, click Edit, and add %SPARK_HOME%\bin.

Note: On Windows, Spark also needs winutils.exe for Hadoop. Place it in a bin folder and point HADOOP_HOME to the parent folder if Spark shows Hadoop errors.

Step 4: Check the installation

Open a new Command Prompt and run:

spark-shell --version

The Spark version banner appears.

Quick recap

Install Java first and check it with java --version.
Extract Spark to a short path like C:\spark.
Set SPARK_HOME and add %SPARK_HOME%\bin to Path.
Run spark-shell --version in a new window to confirm.