Databricks DevBox¶
Welcome to Databricks DevBox - a powerful web-based development environment management platform that brings VS Code to your browser, powered by code-server and optimized for Databricks workflows.
What is Databricks DevBox?¶
Databricks DevBox is a Go-based server application that manages multiple code-server instances, allowing you to create isolated development environments with custom extensions, configurations, and workspace setups. It's designed to be deployed as a Databricks Lakehouse App, providing developers with instant access to cloud-based IDEs directly within their Databricks workspace.
Key Features¶
Instant Development Environments¶
Create and launch isolated code-server instances in seconds with pre-configured extension groups for Python, Jupyter, Databricks, and more.
AI-Powered Coding Assistants¶
Built-in support for multiple vibe coding tools:
- Claude Code: Anthropic's AI coding assistant
- Claude Code Router (CCR): Proxy for Databricks-hosted AI models
- OpenAI Codex: GitHub Copilot-style code completion
- Google Gemini: Google's generative AI assistant
Workspace Templates¶
Quick-start templates for common scenarios:
- Databricks Workshops
- Python Data Science projects
- Machine Learning experiments
- Custom GitHub repository cloning
Real-time Monitoring¶
Track server health, CPU usage, memory consumption, and uptime for all running instances through a modern React-based UI.
Enterprise Security¶
- Databricks SDK authentication
- Auto-generated tokens with configurable expiry
- Unity Catalog integration
- Isolated workspaces per instance
Architecture Overview¶
graph TB
User[User Browser] --> WebUI[React Web UI]
WebUI --> GoServer[Go Server :8000]
GoServer --> CS1[code-server :8010]
GoServer --> CS2[code-server :8011]
GoServer --> CS3[code-server :...]
CS1 --> WS1[Workspace 1]
CS2 --> WS2[Workspace 2]
CS3 --> WS3[Workspace N]
GoServer --> DB[(servers.json)]
GoServer --> Logs[Log Manager]
The system consists of three main components:
- Go Server (
databricks_devbox_go/): Core server managing code-server processes, routing, and lifecycle - Python Wrapper (
app/): Handles Databricks App deployment, token management, and vibe coding tools setup - Web UI (
web_ui/): React-based interface for managing servers
Quick Start¶
As a Databricks App¶
- Copy the
app/folder to your Databricks workspace - Create a Databricks App from the workspace
- The app automatically downloads required binaries (Go server, code-server, Databricks CLI)
- Access the UI through your Databricks App URL
Local Development¶
# Clone the repository
git clone https://github.com/stikkireddy/databricks-devbox.git
cd databricks-devbox
# Install dependencies
make install
# Run the server
make run
# Access at http://localhost:8000
Use Cases¶
1. Training & Workshops¶
Deploy pre-configured development environments for Databricks training sessions with all required extensions and sample code.
2. Remote Development¶
Provide team members with consistent, cloud-based IDEs that work from any device with a browser.
3. AI-Assisted Coding¶
Leverage multiple AI coding assistants (Claude Code, Codex, Gemini) within your development environment.
4. VS Code-Like Development¶
Full-featured IDE environment with extensions, terminal, debugging, and Git integration - not limited to notebooks.
What Makes It Different?¶
Not VS Code Desktop¶
Databricks DevBox uses code-server, which is an open-source fork of VS Code that runs in the browser. This means:
- Web-based: Access from any device
- No installation: Works immediately in browser
- Remote execution: Code runs on server
- Different marketplace: Uses Open VSX instead of Microsoft's marketplace
Learn more about code-server →
Multi-Instance Management¶
Unlike running a single code-server instance, Databricks DevBox:
- Manages multiple isolated instances
- Provides dynamic port allocation
- Includes health monitoring and auto-restart
- Offers a web UI for lifecycle management
Next Steps¶
-
Deploy to Databricks or run locally
-
Customize extension groups and settings
-
Set up AI-powered development tools
-
Integrate with REST API
Community & Support¶
- GitHub: stikkireddy/databricks-devbox
- Issues: Report bugs or request features
- Contributions: Pull requests welcome!
License¶
MIT License - see LICENSE file for details.