Skip to content

ARCTraining/hpc2-software

Repository files navigation

Project Status: WIP – Initial development is in progress, but there has not yet been a stable, usable release suitable for the public.

👷‍♀️ HPC2: Installing and Managing Applications on the HPC 👷‍♂️

This project is a work in progress and is not yet in a stable state.

The home for all course notes for HPC2: Installing and managing application on the HPC Research Computing course.

Lesson Format

We are currently basing our lessons on the Software Carpentry format (e.g. https://swcarpentry.github.io/shell-novice/) which sets a good example of language and structure to follow. Currently these lessons are 25 minutes long (to facilitate a 10 minute break each hour). The current lesson format to follow as a guide:

  • Overview (questions/objectives)
  • Intro paragraph
  • Worked Examples
  • Exercises
  • Key points

Contributing to this project

Working with this project locally

You will need to install the conda package management tool before you can get this project working locally. You can use the environment.yml file included to create a conda environment that contains all the dependencies required to get started.

$ git clone https://github.com/ARCTraining/template-jb-docs.git

$ conda env create -f environment.yml

To build the html content locally you can use the jupyter-book command line tool:

# navigate to the repository root
$ cd template-jb-docs
# sometimes worth running jupyter-book clean book/ to remove old files
$ jupyter-book build book/

Windows

Jupyterbook now supports Windows so the above steps can also be used on a Windows terminal.

Set up a development environment using Vagrant

To aid with this we have created a Vagrantfile that can allow Windows users who have a virtualisation provider installed (such as VirtualBox) and Vagrant installed to create a headless virtual Linux machine that will build the jupyter book. You can do this with the following steps once you've installed a virtualisation provider and vagrant:

# within git-bash or powershell
$ cd template-jb-docs
$ vagrant up

# to rebuild the site after changes with the vagrant box running
$ vagrant reload --provision

# don't forget to destroy the box when you're done
$ vagrant destroy

This will build the jupyter-book html files on your Windows file system (by navigating via /vagrant) so your local build will still persist after you've destroyed your vagrant box.