Demo for the presentation to the meetup datascience in Finance.
To set up the POC :
- You'll need the following env vars in a file .env at the root of the repo
DOCKER_NETWORK_NAME=jupyter
DOCKER_NOTEBOOK_IMAGE=custom_notebook
DOCKER_NOTEBOOK_DIR=/work
DOCKER_SPAWN_CMD=start-singleuser.sh
DATA_VOLUME_HOST=jupyterhub-data
DATA_VOLUME_CONTAINER=/data
GITHUB_CLIENT_ID=
GITHUB_CLIENT_SECRET=
OAUTH_CALLBACK_URL=https://0.0.0.0/hub/oauth_callback
HOST_ARTIFACTORY_HOME=/home/<youruser>/Dev/artifactory_home
You can find GitHub client ID and secret by following registering your application here The host artifactory home is the volume were Artifactory data will be mounted, you can pick any folder really
- Build the custom notebook
cd custom_notebook
docker build -t custom_notebook .
- Launch Artifactory + the Jupyter Server
cd <root_of_the_repo>
docker-compose up
-
Create a remote repository on artifactory called global-pypi. User password is admin password. URL is 0.0.0.0:8081
-
You're all set, you can connect to JupyterHub ! https://0.0.0.0