JuBo

A suite of tools for building and deploying Bokeh apps from Jupyter notebooks using JupyterLabs. It's an experiment to explore a way of quickly and easily creating sharable services from withing Jupyter Lab.

Three main components parts:

• JuBo Lab - A JupyterLab extension. Under the src directory.
• A Module under the jubo directory which consists of:
  • Jubo Server - jubo_server_ext.py - A JupyterServer extension that works with the lab extension to convert and server Bokeh apps from notebooks and 'deploy' by pushing into an S3 bucket.
  • Jubo deployer - proxy.py - A Tornado server that is expecting to run on a Kubernetes cluster. This will spin up, deploy, proxy to and tidy up the bokeh apps built. There is a Helm Chart to deploy this in the kube dir.

A demo

Assumptions/external setup requirements

Installation

pip install git+https://github.com/met-office-lab/jubo.git --user
jupyter serverextension enable --py jubo --user

Development

Jubo Lab and Jubo Server

For a development install (requires npm version 4 or later), do the following in the repository directory:

npm install
npm run build
jupyter labextension link .

If developing the lab extention run in watchmode and re-run npm run build on changes:

jupyter lab --NotebookApp.nbserver_extensions="{'jubo.jubo_server_ext':True}" --notebook-dir=example_notebooks --watch

If working on server-side drop the watch and quit and restart to see changes. This seems quicker than rebuilding JupyterLab constantly.

jupyter lab --NotebookApp.nbserver_extensions="{'jubo.jubo_server_ext':True}" --notebook-dir=example_notebooks

Jubo Deployer

Set up helm with your kubernetes cluster. Then to install in the namespace juboapps as the release juboapps

helm install --namespace=juboapps --name=juboapps .