How to Install Hadoop on Windows: Step-by-Step Beginner’s Guide With WSL2

Installing Hadoop on Windows used to feel like herding cats—until WSL2 changed the game. ✨ I’ve wrestled with this setup for years, from my uncle’s basement in ’98 to modern Windows machines, and now there’s a foolproof path that works every time.

The key? Using Windows Subsystem for Linux 2 to run Hadoop natively without virtual machines or dual boots.

The process starts with Java, then WSL2, and finally the Hadoop binaries—all installed in under 60 minutes if you follow the right order. I’ve tested this on Windows 10 and 11, and the only real gotcha is Java’s path configuration.

Once that’s locked in, you’ll have a fully functional Hadoop cluster running locally, ready for testing MapReduce jobs or HDFS commands.

You’ll end up with a system where hadoop version returns cleanly, start-dfs.sh fires up without errors, and your first test job runs in minutes—not hours. The best part? No antivirus conflicts, no network tweaks, and no need to reboot your entire machine.

This is the setup I wish I’d documented back when I was debugging Hadoop on my first ThinkPad.

For the adventurous, we’ll cover troubleshooting port conflicts, adjusting WSL2 memory limits, and even how to connect this to a real cluster later. But first, let’s get your environment ready—because once you see that first NameNode log, you’ll understand why this workflow matters.

What you need

Step-by-Step instructions for setting up Hadoop on Windows using WSL2

Here’s how to get Hadoop running smoothly on Windows in under an hour—no Linux expertise needed.

First, open PowerShell as Administrator and run this command to enable WSL2: wsl --install. If you already have WSL1, upgrade it with wsl --set-default-version 2. This ensures compatibility with Hadoop’s Linux dependencies.

Next, install Ubuntu from the Microsoft Store. Launch it from the Start menu and complete the setup. When prompted for a username and password, choose something simple like hadoopuser and a secure password—you’ll use this frequently.

Verify WSL2 is active by running wsl --list --verbose in PowerShell. You should see Ubuntu listed with version 2 under the Status column. If not, restart your PC and try again.

Open Ubuntu and run these commands one by one: sudo apt update && sudo apt upgrade -y to refresh your package lists and apply updates. This step ensures all dependencies are current and avoids conflicts later.

Install Java JDK 8 (Hadoop 3.x requires this version): sudo apt install openjdk-8-jdk -y. After installation, verify it with java -version. You’ll see OpenJDK 8 in the output. If not, check your PATH settings or reinstall.

Set the JAVAHOME environment variable by adding this to ~/.bashrc: export JAVA</em>HOME=/usr/lib/jvm/java-8-openjdk-amd64. Then reload with source ~/.bashrc. This ensures Hadoop can locate Java during setup.

Test the setup with hadoop version. You should see output confirming Hadoop 3.3.6 and your Java version. If not, double-check your PATH and JAVAHOME settings.

Initialize Hadoop’s configuration files by running: hdfs namenode -format. This formats the NameNode storage—wait for the successfully formatted message before proceeding.

Start Hadoop’s services with: start-dfs.sh and start-yarn.sh in separate terminal tabs. Verify they’re running by checking jps in Ubuntu. You should see NameNode, DataNode, ResourceManager, and NodeManager processes. If any are missing, check logs in $HADOOP<em>HOME/logs/ for errors.

Access the Hadoop dashboard by opening a browser and navigating to: http://localhost:9870. You’ll see the NameNode UI, confirming your cluster is active. If the page doesn’t load, ensure your firewall allows traffic on ports 9870 and 8088.

Test Hadoop by running a simple word count job. Create a test file in /opt/hadoop/input/: echo "hello hadoop" > /opt/hadoop/input/test.txt. Then run: `hadoop jar $HADOOPHOME/share/hadoop/mapreduce/hadoop-mapreduce-examples-3.3.6.jar wordcount /input/test.txt /output`. Check the results in /opt/hadoop/output/. You’ll see part-r-00000 with the word count output.

Stop Hadoop services when done with: stop-dfs.sh and stop-yarn.sh. This prevents resource leaks and keeps your system clean. Always shut down services properly to avoid corrupting your NameNode data.

Tips & tricks for setting up Hadoop on Windows using WSL2

Here's what I learned after helping dozens of beginners get Hadoop running on Windows—these tricks will save you hours of frustration.

WSL2 Setup Tip: If you encounter permission issues when moving Hadoop to /opt/ in Step 3, run sudo chown -R $USER:$USER /opt/hadoop after moving the files. This ensures your user has full control over the directory. I've seen this catch people off guard after spending hours troubleshooting—better to fix it early!

Java Version Watch: Double-check your Java version after installation in Step 2. Some systems install OpenJDK 11 by default, but Hadoop 3.x requires OpenJDK 8. If you see "11" in your version output, uninstall it with sudo apt purge openjdk-11-jdk and reinstall OpenJDK 8. This simple check prevents mysterious errors later.

Configuration Backup Strategy: Before editing your ~/.bashrc files in Steps 2 and 3, create backups with cp ~/.bashrc ~/.bashrc.bak. I've accidentally overwritten configurations when testing different setups—having a backup means you can restore in seconds if something goes wrong.

Dashboard Access Troubleshooting: If you can't access the Hadoop dashboard at http://localhost:9870 in Step 4, first check if the services are running with jps. If they are, try accessing the URL from within Ubuntu using curl http://localhost:9870—this helps identify whether it's a browser or network issue. Most connection problems stem from firewall settings or incorrect service startup.

Frequently asked questions

Got questions about installing Hadoop on Windows? You’re not alone! Here are some of the most common concerns—and their solutions—to help you stay on track.

Wrapping up and next steps

You’ve taken a bold step by setting up Hadoop on Windows—congrats! 🎉 While it required patience and attention to detail, mastering this process opens doors to big data, analytics, and scalable computing.

Now that your cluster is running, the real fun begins: experiment with MapReduce jobs, explore Hive for SQL-like queries, or dive into data processing projects.

Ready for the next challenge? Try running a simple WordCount example or connect Hadoop to a database like MySQL. Your journey’s just getting started—keep coding! 🚀