FastAPI Deployment to Azure Functions (Learning Project)

This repository demonstrates how to deploy a FastAPI application to Azure Functions using the v2 model with a function_app.py file. The goal of this project is to showcase the deployment process of a basic FastAPI app using Azure’s serverless computing capabilities.

🔗 GitHub Repository

GitHub Repository Link

📋Table of Contents

• About
• Prerequisites
• FastAPI Application Setup
• Creating an Azure Function
• Deploying FastAPI to Azure Functions
• Testing the Deployment
• What I Learned
• Conclusion
• References

🌟About

This is a learning project aimed at showcasing the deployment process of a FastAPI application to Azure Functions. I am working on FastAPI-related tasks at Mila AI, and this project helps me apply and demonstrate my knowledge of deploying cloud applications, focusing on the serverless approach using Azure.

---

🛠Prerequisites

Before starting, ensure you have the following:

• Azure CLI: Install Azure CLI.
• Azure Functions Core Tools: For running functions locally. Install here.
• FastAPI & Uvicorn: Install using:

  pip install fastapi uvicorn

---

FastAPI Application Setup

Once the FastAPI application code is set up with a basic structure, you'll need to ensure that it can handle HTTP requests when deployed as an Azure Function. The setup includes a simple FastAPI app with one or two routes to demonstrate the deployment.

We will use the v2 model of Azure Functions, which structures the project using a function_app.py file, eliminating the need for a function.json file.

---

Creating an Azure Function

Next, we’ll initialize the Azure Function and set up the HTTP trigger to route requests to our FastAPI app.

1. Initialize a new Azure Function in your project directory:

   func init --worker-runtime python

2. Create an HTTP-triggered function:

   func new --name HttpTrigger --template "HTTP trigger" --authlevel "anonymous"

3. Modify function_app.py to route requests to FastAPI app (as given in codebase).

---

🚀Deploying FastAPI to Azure Functions

Deploying the FastAPI app to Azure Functions involves the following steps:

1. Login to Azure via CLI:

   az login

2. Deploy the App:

   • First, create a ZIP file of your function app:

     zip -r function.zip .

   • Then, deploy the ZIP file to your Azure Function App:

     az functionapp deployment source config-zip --name <YOUR_FUNCTION_APP_NAME> --resource-group <YOUR_RESOURCE_GROUP_NAME> --src function.zip

---

Testing the Deployment

Once deployed, you can test your FastAPI app by accessing the provided function URL.

1. Get the URL: The Azure CLI will output the URL for function app after deployment. It will look something like:

   https://<YOUR_FUNCTION_APP_NAME>.azurewebsites.net/HttpTrigger
   

2. Test the API: Use curl or Postman to send a request to your FastAPI endpoint:

   curl https://<YOUR_FUNCTION_APP_NAME>.azurewebsites.net/HttpTrigger

   You should see a JSON response similar to:

   {
       "message": "Hello from FastAPI deployed on Azure Functions!"
   }

---

🧠What I Learned

While working on this project, I gained hands-on experience with:

• Deploying FastAPI apps to Azure Functions using the serverless model.
• Understanding how WSGI middleware works to integrate FastAPI with Azure Functions.
• Using Azure CLI for managing resources, creating Function Apps, and deploying applications.
• Configuring FastAPI to work in a serverless environment with automatic scaling based on demand.

---

🎯Conclusion

This project serves as a demonstration of deploying a simple FastAPI application to Azure Functions using the v2 model. While this project focuses on the deployment process, it can be extended into a full-featured API in the future.

Feel free to explore, contribute, or provide feedback on this repository!

---

🔗References

This project was built based on the official Microsoft Azure documentation. You can learn more about deploying Python apps on Azure Functions from the following resources:

• Azure Functions Python Developer Guide

---