JupyterHub Data Science Server on Ubuntu 24.04 LTS

Overview

JupyterHub Multi-User Data Science Server on Ubuntu 24.04 LTS is a preconfigured virtual machine image designed to help organizations quickly deploy a centralized, browser-based data science and notebook environment.

The VM includes JupyterHub for multi-user notebook access, JupyterLab as the default user interface, Nginx as the public reverse proxy, and systemd for service management. It is designed for teams that need a shared, secure, and repeatable data science workspace without manually configuring JupyterHub from scratch.

This image is suitable for data science teams, analytics teams, education environments, internal engineering teams, AI/ML labs, and organizations that need a multi-user Python notebook server running on Ubuntu 24.04 LTS.

Key Capabilities

• Preconfigured JupyterHub deployment
• JupyterLab enabled as the default user interface
• Runs as a managed systemd service
• Uses Nginx as the front-end reverse proxy
• Supports multiple Linux users
• Uses Linux PAM authentication
• Supports group-based JupyterHub access
• Automatically allows Linux sudo users to access JupyterHub as admins
• Includes a global Python 3 Data Science kernel
• Includes common data science and machine learning libraries
• Optional support for GPU workloads with PyTorch CUDA packages
• Designed for repeatable VM image deployment
• Built on Ubuntu 24.04 LTS

Included Software

The VM includes the following core components:

Included Python Data Science Libraries

The shared global data science kernel includes:

• numpy
• pandas
• matplotlib
• seaborn
• scikit-learn
• scipy
• statsmodels

The registered shared kernel is displayed in JupyterLab as:

Python 3 (Data Science)

Authentication and User Access

This VM uses local Linux user accounts for authentication.

Users are allowed to access JupyterHub if they are members of either:

jupyterhub-users

or:

sudo

Members of the Linux sudo group are automatically granted JupyterHub administrator access.

Example User Creation

sudo adduser analyst1
sudo usermod -aG jupyterhub-users analyst1
sudo systemctl restart jupyterhub

Example Admin Creation

sudo adduser adminuser
sudo usermod -aG sudo adminuser
sudo systemctl restart jupyterhub

Network Architecture

Nginx listens publicly on port 80 and proxies traffic to JupyterHub.

Typical flow:

User Browser → Nginx → JupyterHub → User Notebook Server

This allows Nginx to act as the public-facing web service while JupyterHub remains internal to the VM.

Default Access URL

After deployment:

http://<SERVER-IP>/hub/admin

Recommended Use Cases

This VM is ideal for:

• Team-based data science environments
• Internal analytics workbenches
• AI/ML experimentation
• University or classroom notebook environments
• Shared Python development environments