How to Set Up Databricks Free Edition for Databricks Sword | Databricks Sword Blog | Databricks Sword

How to Set Up Databricks Free Edition for Databricks Sword

Databricks Sword works entirely in simulated mode — you never need a real Databricks workspace. But if you want to run your code against a real Spark cluster, Databricks offers a free Community Edition that's perfect for learning.

This guide walks you through setup in under 10 minutes.

What Is Databricks Community Edition?

Databricks Community Edition is a free, limited version of the full Databricks platform. It gives you:

A single-node Spark cluster (driver only, no workers)
Notebooks for writing PySpark, SQL, Scala, and R
Delta Lake support (ACID transactions, time travel)
DBFS (Databricks File System) for storing data
MLflow for experiment tracking

Limitations

No multi-node clusters (single driver only)
No Jobs/Workflows scheduler
No Unity Catalog
No SQL Warehouses
Cluster auto-terminates after 2 hours of inactivity
Community support only (no SLA)

For Databricks Sword missions, the Community Edition covers all B-rank and most A-rank content. S-rank missions that require Unity Catalog or Workflows will remain in simulated mode.

Step 1: Sign Up

Navigate to community.cloud.databricks.com
Click Sign Up (or Get Started — Free)
Fill in:
- First name, last name
- Email address (use a personal email — corporate emails may redirect to a trial)
- Company (enter "Learning" or your actual company)
Check the terms of service checkbox
Click Sign Up
Verify your email address by clicking the link sent to your inbox

Tip: If you're redirected to a 14-day trial page instead of Community Edition, look for a small link that says "Get started with Community Edition" near the bottom.

Step 2: Create Your Workspace

After verifying your email:

You'll be prompted to choose a cloud provider — select Community Edition (not AWS/Azure/GCP)
Wait 1-2 minutes for your workspace to provision
You'll land on the Databricks Workspace home page

Your workspace URL will look like: https://community.cloud.databricks.com

Explore the Interface

Workspace — your notebooks and files
Repos — Git integration
Data — browse tables and DBFS
Compute — manage your cluster
Machine Learning — MLflow experiments

Step 3: Start Your Cluster

Before running any code, you need an active cluster:

Click Compute in the left sidebar
Click Create Cluster (or your existing cluster if one exists)
Give it a name like databricks-sword-cluster
Leave all defaults (single node, latest runtime)
Click Create Cluster
Wait 3-5 minutes for the cluster to start (status changes from "Pending" to "Running")

Note: The cluster auto-terminates after 2 hours of inactivity. You can restart it anytime from the Compute page.

Step 4: Generate a Personal Access Token (PAT)

To connect your Databricks workspace to Databricks Sword, you need a PAT:

Click your profile icon (top-right corner)
Select Settings
Navigate to Developer → Access Tokens
Click Generate New Token
Enter a description: Databricks Sword
Set expiration: 90 days (or your preference)
Click Generate
Copy the token immediately — you won't be able to see it again!

Store the token somewhere safe (password manager recommended).

Step 5: Connect to Databricks Sword

Open Databricks Sword at your deployed URL
Navigate to Profile → Settings
Find the Databricks Connection section
Enter:
- Workspace URL: https://community.cloud.databricks.com
- Personal Access Token: paste your PAT
Click Verify Connection
If successful, you'll see a green checkmark
Click Save

Step 6: Enable Real Mode on Missions

Now that you're connected:

Open any mission
Look for the Mode Toggle at the top of the stage player
Switch from Simulated to Real Databricks
Your code will now execute against your real Databricks cluster

Important: Not all stages support real mode. Drag-drop and quiz stages are always simulated. Code challenges (fill-blank, free-text) will execute against your workspace.

What's Next?

With your workspace connected, you can:

Run PySpark code against a real Spark cluster
Create Delta tables and query them with time travel
Track MLflow experiments in your workspace
Compare your results with the simulated output

Start with a B-rank mission like Lakehouse Fundamentals to test your connection.

Troubleshooting

"PAT expired" Error

Your Personal Access Token has a set expiration date. Generate a new one:

Go to Settings → Developer → Access Tokens
Generate a new token
Update it in Databricks Sword Profile → Settings

"Cluster not running" Error

Your cluster auto-terminates after 2 hours of inactivity:

Go to Compute in your Databricks workspace
Find your cluster and click Start
Wait for it to reach "Running" state
Retry the mission

"Connection refused" Error

Check that:

Your workspace URL is exactly https://community.cloud.databricks.com
There are no trailing slashes
Your PAT is correct and hasn't expired
Your cluster is running

"Redirected to trial" During Signup

If you keep getting redirected to the 14-day trial:

Try using a different email address (personal Gmail, etc.)
Look for "Community Edition" link at the bottom of the trial page
Clear your browser cookies and try again
Use an incognito/private window

Cluster Takes Too Long to Start

Community Edition clusters can take 3-7 minutes to start. If it takes longer than 10 minutes:

Cancel the cluster creation
Create a new cluster
Try during off-peak hours (early morning or late evening UTC)

That's it! You're now ready to level up your Databricks skills with real cluster execution. Remember — simulated mode works perfectly for all missions. Real mode is optional but recommended for the full experience.

Happy forging! ⚔️