JupyterHub: A multi-user server for Jupyter notebooks

JupyterHub is a multi-user server that manages and proxies multiple instances of the single-user IPython Jupyter notebook server.

Three actors:

• multi-user Hub (tornado process)
• configurable http proxy (node-http-proxy)
• multiple single-user IPython notebook servers (Python/IPython/tornado)

Basic principles:

• Hub spawns proxy
• Proxy forwards ~all requests to hub by default
• Hub handles login, and spawns single-user servers on demand
• Hub configures proxy to forward url prefixes to single-user servers

Dependencies

JupyterHub requires IPython >= 3.0 (current master) and Python >= 3.3.

You will need nodejs/npm, which you can get from your package manager:

sudo apt-get install npm nodejs-legacy

(The nodejs-legacy package installs the node executable, which is required for npm to work on Debian/Ubuntu at this point)

Then install javascript dependencies:

sudo npm install -g configurable-http-proxy

Installation

Then you can install the Python package by doing:

pip install -r requirements.txt
pip install .

If you plan to run notebook servers locally, you may also need to install the IPython notebook:

pip install "ipython[notebook]"

This will fetch client-side javascript dependencies and compile CSS, and install these files to sys.prefix/share/jupyter, as well as install any Python dependencies.

Development install

For a development install:

pip install -r dev-requirements.txt
pip install -e .

In which case you may need to manually update javascript and css after some updates, with:

python setup.py js    # fetch updated client-side js (changes rarely)
python setup.py css   # recompile CSS from LESS sources

Running the server

To start the server, run the command:

jupyterhub

and then visit http://localhost:8000, and sign in with your unix credentials.

If you want multiple users to be able to sign into the server, you will need to run the jupyterhub command as a privileged user, such as root. The wiki describes how to run the server as a less privileged user, which requires more configuration of the system.

Some examples

generate a default config file:

jupyterhub --generate-config

spawn the server on 10.0.1.2:443 with https:

jupyterhub --ip 10.0.1.2 --port 443 --ssl-key my_ssl.key --ssl-cert my_ssl.cert

The authentication and process spawning mechanisms can be replaced, which should allow plugging into a variety of authentication or process control environments. Some examples, meant as illustration and testing of this concept:

• Using GitHub OAuth instead of PAM with OAuthenticator
• Spawning single-user servers with docker, using the DockerSpawner