Dockerizing Angular 2, Flask, NodeJS, and MongoDB With Docker-Compose and Docker-Machine

Goal:

The goal of this repository is to have a quickly sharable, deployable docker setup for Python and Node web services adhering to best-practices. Nginx serves the static content and provides a reverse proxy to Python/Flask and NodeJS/Express web service apis. Since Nginx, Python, and NodeJS are in separate (linked) docker containers, each service has a separate IP from the host docker-machine. Due to this, nginx reverse-proxies requests over specific TCP ports between containers, as opposed to using a unix socket implementation. A benefit of this linked container implementation is a smaller footprint than a docker server image, and that proxies do not go through the externally-exposed docker-machine host.

Stack:

• Docker v1.11.1
• Docker Compose v1.7.1
• Docker Machine v0.7.0
• Nginx
• MongoDB
• Python API -> Flask / Flask-RESTful behind Nginx reverse-proxy for data science (Python 2.7 for PyLucene integration)
• uWSGI -> Written in c, use of this uWSGI server runs the Python/Flask application very efficiently.
• NodeJS API -> Express behind Nginx reverse-proxy for high-performance http requests and json handling
• Angular 2 -> SPA on a cutting-edge framework with native lazy loading

Getting up and running:

• Install Docker Toolbox - https://docs.docker.com/engine/getstarted/step_one/

• Run the following from a terminal within this project root:

  npm install

  npm run dev

Verify proper docker build from the browser:

• Retrieve docker-machine ip address from the terminal/cmd prompt:

  docker-machine ip

• Static assets served from Nginx:

  • http://docker-machine-ip/

• Flask API REST end-point:

  • http://docker-machine-ip/flask-api/user
  • http://docker-machine-ip/flask-api/users

• NodeJS API REST end-point:

  • http://docker-machine-ip/node-api/user
  • http://docker-machine-ip/node-api/users

Reminder: "docker-machine-ip" above refers to the host IP of the docker-machine returned from the terminal on this command:

docker-machine ip

Cheers!