About jupyter

levell jupyterhub container definition.

Frameworks used

Jupyterhub
Jupyter
scipy

Docker image details

Jupyterhub

Base image: jupyterhub/jupyterhub
Exposed ports: 8000
Additional installed resources:

Troubleshooting: vim, net-tools, dos2unix
Python libraries: dockerspawner, oauthenticator, jupyterhub-idle-culler

Jupyter notebook

Base image: jupyter/scipy-notebook:33add21fab64 # special version for jupyterhub (https://github.com/jupyter/docker-stacks)
Exposed ports: 8888
Additional installed resources:

Troubleshooting: vim, net-tools, dos2unix
Data science: sklearn, pandas, seaborn

Config

The service configuration can be done in the jupyterhub_config.py which is copied into the docker container at build time.

For the authentication the auth0 client secret has to be stored in as docker secret:

printf "clientsecret_auth0" | docker secret create auth0_jupyterhub_client_secret -

Deployment

General

Service: jupyter
Data Path: /home/worker/levell/jupyter/
Access URL: jupyter.app.levell.ch

Attached Networks

traefik-public - access to reverse proxy
levell_jupyter - attachable overlay network used by the spawned jupyter notebook containers

docker network create --driver=overlay levell_jupyter --attachable

Attached volumes

jupyterdata: storing the jupyter notebooks, for each user a sub directory is created.

Environment variables

DOCKER_JUPYTER_CONTAINER: jupyter/scipy-notebook:33add21fab64 # image the spawned containers are based on DOCKER_NETWORK_NAME: levell_jupyter # name of the network used for the spawned containers CLIENT_SECRET: /run/secrets/auth0_jupyterhub_client_secret # path to docker secret of the auth0 client secret HUB_IP: 0.0.0.0 # binding IP address of the jupyterhub

Authentication

Auth0

jimmylevell/jupyterhub