toolbox

Toolbox is a development container you can customize to publish solid, stable development environments for software and systems engineers. Based on the continuumio/miniconda3 image, it's the Docker equivalent to virtualenv.

Umm... why?

We always deploy and consider the data scientist or engineer's workstation the single most important deployment in a project. This lets us publish consistent, stable development environments for data scientists and engineers that we can throw away and rebuild due to personal preferences or requirement changes.

Nuts and Bolts

Dockerfile - container definition file

Installs container system packages, Docker, and Python libraries. This is where you would install a database, web server, or build tools.

bin/entrypoint.sh - container's executable

Runtime configuration. This is where you would install project dependencies and shell configuration i.e. Python libraries, bash settings, shell scripts

bin/build.sh - builds and runs the container

I don't want to run the container on my workstation

All you need is docker build -t idahodata/toolbox .

I would love to run the container on my workstation

The following files are mapped inside your Toolbox container:

workstation	container
$HOME/.gitconfig	/home/docker/.gitconfig
$HOME/.ssh	/home/docker/.ssh
/var/run/docker.sock	/var/run/docker.sock

The first are required so git works properly inside Toolbox.

Mapping /var/run/docker.sock lets us control the workstation's Docker host inside the container so we can build and use other containers. A future feature will have a local DNS server so we can use fully-qualified domain names in development which, at the least, makes managing development HTTPS certificates much easier.

Howto

We use Toolbox for vcardz development.

1. clone Toolbox and create branch for your project

git clone git@github.com:IdahoDataEngineers/toolbox.git
cd toolbox
git checkout -b vcardz

1. clone your project

git checkout git@github.com:IdahoDataEngineers/vcardz.git

1. customize Toolbox We modified bin/entrypoint.sh to install vcardz as an editable Python library

43: # install vcardz Python module
44: pip install -e vcardz/

1. Run the container

bin/build.sh

and you'll see

✔  [docker@toolbox]  [vcardz|✔]
18:46 $