Docker-compose FastAPI + POSTGRES recap

Set-up

Go to the root of your coding folder and clone the repo:

git clone [email protected]:Michiel-DK/postgres_fastapi_recap.git

Then run below to export env variables:

direnv allow .
direnv reload

Get the data

This example dataset consists of multiple .csv files containig airbnb data in multiple cities.

Run below to merge the datasets + filter out a few columns:

make get_data

Build your images

Build both containers by running below in the terminal:

docker compose build

This will use the docker-compose.yml to create the two images:

• airbnb_api is a python image containg the code to run the api
• airbnb_postgres is a postgres image which will act as the database

Instantiate your containers

After that you can run below to instantiate two containers from the image:

docker compose up

If you look at the containers in the docker app you will see the directory postgres_fastapi_recap which contains:

• airbnb_api_c which runs on http://localhost:8000/
• airbnb_postgres_c which runs on http://localhost:5432/ (note that here you won't able to do much with the browser)

Note that when you run it for the first time it will run setup_tables.sql in the postgres container.

This script does below in order:

1. Creates a database called airbnb
2. Moves into that newly created database
3. Creates an empty table called cities to populate later
4. Populates the database with our merged.csv
5. Deletes the csv from the image

Also have a look at the Volumes tab on the docker-app. You'll see we've created a postgres_fastapi_recap_pg_data volume.

This will allow us to persist the database through different instantiations of the containers. Deleting this will delete all your data so watch out!

Inspect the API

You can then test the api on http://localhost:8000/docs and have a look at the different fastapi related .py files.

Essentially:

• database.py connects the api to the database
• models.py structures the databse content in a Python class
• crud.py defines an operation on the database - here extract all from database based on variable city
• schemas.py allows you to create different response models based on the models
• api.py defines the endpoints of the api