DarthDippy/docker-scaffolding

Template for Docker-based projects

Template Project

This is a project template that creates a Docker environment with useful containers for the Data For Democracy project.

Docker is a technology that allows you to package up all of the dependencies for a project in a portable way. Docker uses "containers", which are like virtual environments, but extending to the whole OS, not just Python packages.

This template sets up a Node webapp server, Flask webapp server, a Conda environment running Jupyter, and Postgres and MongoDB databases that they can talk to. The idea is that you'll delete the pieces that you don't intend to use (remove them from docker-compose.yml and delete the directories), and customize the ones you do.

To get started:

1. Install Docker: https://www.docker.com/products/overview

   If possible, use the native docker setup for your platform:

   • Docker for Windows
   • Docker for Mac
   • Linux supports docker without virtual machines

   If you need to use an earlier version...

   On Windows:

   • Get the docker toolbox for windows 10 home or earlier: https://docs.docker.com/toolbox/toolbox_install_windows/
   • Make sure virtualization is enabled (enter BIOS menu, adjust as needed)
   • Create a docker machine. A good name for the machine is the name of the github repo you'll be using it with, eg. "internal-displacement". For windows 10 home and earlier use docker-machine create --driver virtualbox MACHINENAME Use other drivers as necessary (https://docs.docker.com/engine/getstarted-voting-app/node-setup/)
   • Setup the machine's environment variables: docker-machine env MACHINENAME
   • Finally, activate the machine. The last statement should tell you how but in powershell: docker-machine env MACHINENAME | invoke-expression

   On MacOS

   • Install VirtualBox: https://www.virtualbox.org/
   • Create a docker machine. A good name for the machine is the name of the github repo you'll be using it with, eg. "internal-displacement". docker-machine create MACHINENAME
   • Set up your environment so that the docker commands can access this docker machine: docker-machine env MACHINENAME

2. In this directory, run docker-compose up This will take a while the first time you do it, because it's downloading all of the dependencies for all of the subparts of the project. It should be much faster after the first time.

3. In the output, you should see lines like these, telling you where to connect:

conda_1     |     Copy/paste this URL into your browser when you connect for the first time,
conda_1     |     to login with a token:
conda_1     |         http://localhost:5002/?token=9c010637753fdf9646eb75a3d6b36b0a5e5464e5b5d32df3

flask_1     |  * Running on http://0.0.0.0:5001/ (Press CTRL+C to quit)

node_1      | Running on http://localhost:5003

4. To stop all of the running services, use CTRL+C.

5. Data stored in the database will be preserved from run to run.

6. Code put in source/python will be available to both Conda and the Flask server

7. If you add something to requirements.txt or packages.json use docker-compose up --build to pick up the new dependencies