Introduction

This repo holds the source code for the article: API Service with FastAPI + AWS Lambda + API Gateway and Make it Work. It consists of four main components

1. The FastAPI app inside the app/ folder
2. The Dockerfile to build an image of the FastAPI app
3. scripts/deploy.sh to automatically spin up the necessary AWS resources (AWS Lambda, API Gateway, IAM role, etc.) and deploy the FastAPI app
4. scripts/teardown.sh to automatically tear down the AWS resources needed to run the API service.

Run the FastAPI App Locally

1. Create a virtual environment

   python -m venv venv
   source venv/bin/activate
   

2. Install the dependencies

   pip install -r requirements.txt
   

3. Add .env file to hold the environment variables. The file itself is not committed to the repo per standard practice. Example content is shown below:

   SOME_ENV=SOJ
   OTHER_ENV=Shako
   

4. Run the app inside the app/ folder

   cd app
   uvicorn main:app --reload
   

5. View the OpenAPI documentation at http://127.0.0.1:8000/docs

Deploy the FastAPI App to AWS

The FastAPI app is packaged into a docker image and executed on AWS Lambda.

First create a .env.stage or .env.prod files to hold the environment variables specific to the stage or prod. These two files should NOT be committed.

# example content for .env.stage
SOME_ENV=Anni
OTHER_ENV=Raven

# example content for .env.prod
SOME_ENV=Torch
OTHER_ENV=Griffon

Then run the following script to automate the entire process of building and deploying the FastAPI app to AWS.

IMAGE=fastapi ENV=stage ./scripts/deploy.sh

IMAGE is the docker image name. It is also the name of the Lambda function and the API Gateway. ENV marks the current environment (either stage or prod).

Tear Down the AWS Resources

To tear down all the resources used to run the API service, use the following script

IMAGE=fastapi ENV=stage ./scripts/teardown.sh