AWS Lambda Deployment Pipeline

This repository contains a minimal (3 short files) example of how to use GitHub Actions and Serverless Framework to automate the deployment pipeline for an API backed with AWS Lambda and AWS API Gateway.

Why

AWS includes powerful tools for production deployments but the learning curve to spin up a new service can be steep. This is intended to be used as a starter for other serverless projects, along with the great examples provided in the serverless/examples repo.

AWS Lambda is the current leader in serverless computing and provides the execution environment for user defined functions.

AWS API Gateway can be used to map Lambda function execution to public facing HTTP endpoints and can be configured to perform JWT authentication/authorization, abstracting the auth logic away from application code.

Files

• serverless.yml configures the application to receive events at a given URL.
• handler.py receives the event and returns a response.
• .github/workflows/main.yml configures GitHub Actions to build and deploy the application.

Setup

Clone or copy these files and the corresponding directory structure. None of these files need to be edited to get a working example, since the only configuration necessary is supplying the credentials (instructions below) to allow the GitHub Actions runner to configure the necessary AWS resources.

You'll need an AWS account to deploy the application.

1. Go to the Users page in the AWS Identity and Access Management (IAM) console.
2. Create a new Admin user that will be used by the deployment pipeline.
3. Select this user, then Security credentials > Access keys and Create access key.
4. Add the key credentials to the Secrets settings of your GitHub repository, found at Settings > Secrets using the AWS_ACCESS_KEY_ID and AWS_SECRET_ACCESS_KEY environment variables.

Done! Each new push to master will now build and redeploy the service.