Demos using Jupyterlab

This repo was forked from larryebaum/hc_demos-jupyter. The contents below is a summary of a presentation delivered by Larry Eichenbaum and James Nyika.

Agenda:

• What is Jupyter, Jupyter Lab and the use case
• When to Jupyter
• Existing Content Overview
• Demo
• How the Sausage is made - setting up Jupyter for yourself
• Questions

What is Jupyter / Jupyter Lab ?

• Jupyter Notebook is an open-source web application that allows you to create and share documents that contain live code, equations, visualizations and narrative text
• JupyterLab is a web-based interactive development environment for Jupyter notebooks, code, and data. JupyterLab is the next-generation user interface for Jupyter notebooks that is more modular, allows you to open multiple notebooks at once and provides other useful features.
• Project Jupyter (https://jupyter.org/ ) is where it is hosted and maintained.

The Use Case & Value

• Jupyter Lab can run different kernels. The most common is the python kernel.
• A lesser known kernel is the bash / shell kernel
• 💰: Allows notebooks to be used like a bash shell
• Use Jupyter Lab to execute CLI commands to Vault, Nomad, Consul, or Terraform as part of a demo for a better experience

When to use Jupyter vs Instruqt

Jupyter excels when:

• You need to run a demo scenario with limited time
• You need to focus on telling a story and speaking
• You need to provide your audience with a history of the demo actions in a standard human readable format
• You need a non video recording of your demo
• You need to reuse your demo as a reference document
• You need to customize the experience for the customer

---

Existing Content

• Ready to use demos:
  • http://bit.ly/HC-SE-Demos
  • To be moved into GH hashicorp/solutions_eng sub-org once available.
• Header artwork:
  • http://bit.ly/Jupyter_Headers-James
  • http://bit.ly/Jupyter_Headers-Larry

DEMO

---

Installation / Configuration of Jupyterlab

Official Docs: https://jupyterlab.readthedocs.io/en/stable/

Mac / Homebrew

brew update && brew doctor
brew install python3
brew install jupyterlab
pip3 install --upgrade pip wheel setuptools
#pip3 install ipykernel #installed with jupyterlab
pip3 install bash_kernel
python3 -m bash_kernel.install

For a more custom environment

brew upgrade
brew install pyenv
pyenv versions
echo 'eval "$(pyenv init -)"' >> ~/.bash_profile

Additional Reference: https://medium.com/@blessedmarcel1/how-to-install-jupyter-notebook-on-mac-using-homebrew-528c39fd530f

Windows

You need to do more - install Anaconda and configure: https://mycarta.wordpress.com/2019/07/09/from-zero-to-jupyterlab-pro-on-windows-10/

---

Start Jupyter Lab - Mac and Windows

Go to a folder that holds or will hold your Jupyter Notebook. Start up Jupyter Lab

Mac :

If installed from brew

jupyter lab

If installed from pip

jupyter-lab

or

Windows :

Jupyter lab

Clear outputs

find terraform vault -name '*.ipynb' \
    | xargs -P 6 -n 1 jupyter nbconvert --clear-output --inplace

Print to Notebook

Jupyter Notebook to PDF in a few lines | by Cornellius Yudha Wijaya | Towards Data Science

Extras

Install Latex Mac - http://tug.org/mactex/mactex-download.html

pip install -U notebook-as-pdf
# For chromium
pyppeteer-install
# For CLI
jupyter-nbconvert --to PDFviaHTML example.ipynb

Converting to PDF with Latex ala mactex

jupyter-nbconvert --to pdf example.ipynb

Call to Action

• Setup and successfully run one of these demos on your system.
• Give us feedback on your experience and things you would like to see.
• Contribute back some notebooks and support us in making them a little more standardized.

List of Demos

• Terraform
• Vault
• Consul

Pull requests

Jupyter Notebooks include the output of the commands directly into the Notebook.

• Version control will have issues generating a useful diff.
• Clear all output before uploading the notebook to git. There are other ideas around this as well [5,6], but nothing i looked into yet.