JupyterHub

JupyterHub is a multi-user server for Jupyter notebooks, designed to provide shared access to computational environments for groups, such as data science teams, researchers, and classrooms. It allows multiple users to run and manage their own Jupyter notebook sessions simultaneously on a shared infrastructure. JupyterHub supports authentication, user management, and resource allocation, making it ideal for collaborative and large-scale data science projects. With support for customizable environments, JupyterHub is widely used in education, research, and enterprise AI workflows.

The quick deploy app is based on The Littlest JupyterHub

Get started

Go to the apps section in the web console and click either the small, medium or large instance of JupyterHub. This will give you some good default settings but, you can fully customise your deployment at the next step.

VM configuration

A GPU is automatically selected by clicking small, medium or large, however you can choose a different GPU or even multiple GPUs for your deployment. The default disk size is set between 100-200GB which should be enough for most users. However, if you have a very large dataset to deploy on the machine you may need to increase the size.

Having a large number of users may require larger system memory and vCPU allocations. Read this for more information.

Within jupyterhub usage can be monitored using the right hand side pane:

Using JupyterHub

When you deploy the VM you will be shown the VM information page. Find the public IP address and paste it into your browser using http (https is not supported) e.g. http://192.1.3.4.5

Use the username admin and create a password to sign in. The first user to sign in will have an account created for them and become the administrator.

Got to File > Hub Control Panel to open the hub control panel. Click Admin to show the admin view.

Click Add Users to add a list of users, each username should be on a new line.

Users can now sign in and create their own password.

There are more options for creating accounts and authentication here.

Installing packages

Conda, pip and apt packages can be install for all users via the terminal, please follow this guide carefully.

Adding data and sharing it

You can add data to JupyterHub and share it with your users by following this guide