PeopleGetter

High-level usage

1. Start up the backend service containers (a Python API and a Postgres database), and the webserver (nginx)

./start_all.sh

2. Use the live server extension to open frontend/index.html

The start_all.sh script

The start_all.sh script does four things:

1. Runs the db/setup.sh script
2. Dynamically gets the db container's IP address and writes it to api/db_ip_addr so the api container can pick it up
3. Runs the api/start_api.sh script
4. Runs the frontend/start_frontend.sh script

The Database

The db folder contains everything needed to run a containerized Postgres database with a pre-loaded Person table. The setup.sh script will build the image from the Dockerfile, then run it as a background process.

This service needs to be running before the Python API container, because the Python API will need to have access to the database's IP address, which is assigned by Docker when the container starts.

The API

The api folder contains everything you need to run a Flask API (Python). Because Flask is a 3rd-party module, we have to make sure that the container has the necessary dependencies installed (listed in requirements.txt).

The server will listen for requests on port 8000 (this is configurable). When a request comes in to either /all or /<id> (i.e /3, /973, etc.), psycopg2 will create a connection to the database container and fetch results. These results will then be converted to JSON and sent as an HTTP response.

The tricky part with this service is that it can't know what the database container's HOST is until after the database container is running. Because of that, the HOST must be configured dynamically.

That configuration is handled by the db_ip_addr file. the db_ip_addr file is created during the execution of the start_all.sh script from this command:

docker inspect  -f '{{range.NetworkSettings.Networks}}{{.IPAddress}}{{end}}' pg_db > api/db_ip_addr

That's quite a mouthful, but it's just parsing through the results of docker inspect pg_db, which as you can see if you execute it while the database container is running, is a bunch of JSON describing the running container. We just need one little piece of information from that: the IPAddress.

The database container's IP address is then written to the db_ip_addr file. That file is then copied into the api container when it's built. Then when python server.py is executed, the contents of the db_ip_addr are read in and made available to the server.

The frontend

The frontend folder contains everything you need for a web server. It uses the Alpine Nginx image. Configuring a web server can get complicated, but fortunately all the default configuration will serve our purposes for now. All we have to do is copy our html/css/js files to /usr/share/nginx/html/ in the container. This is just how Nginx works, it has nothing to do with Docker.

Visit the site at http://localhost/