This is an example project for Serverless in AWS

We will be using SAM (https://github.com/awslabs/serverless-application-model)

Most services will be using the free tier so costs should be very minimal.

1. Signup for an AWS account
2. Fork this project into your GitHub account
3. Python 3.9
4. Postman
5. Set up a virtual environment
6. Install dependancies
   • boto3
   • aws-dynamodb-parser
   • coverage

1. Run unit tests coverage run --branch --source='.' -m unittest and coverage with coverage report -m --fail-under=100 --omit=*/__init__.py,tests/*,cookie-shop-env/*
2. Run the setup-template.yaml in AWS Cloudformation
   • Give the S3 bucket a globally unique name (line 78)
   • If this fails it is most likely because Cloudformation cannot setup CodeBuild projects without first connecting to Github through OAuth. You should create a dummy CodeBuild project and connect to Github.
   • Explore the created resources in AWS - S3, IAM, CodeBuild, SSM
3. Create a PR in the repo
   • This build will use buildspec-dev.yaml and should create a new Cloudformation stack (dev) and create all the resources for us. Note this will be a seperate stack for each PR.
   • Explore the created resources in AWS - Dynamo, Lambda, API Gateway, etc.
4. Use the API's
   • Go to API Gateway -> Stages -> Prod and get the Invoke URL
   • Use swagger to figure out what to send the API, then postman to call it
   • Try the /cookies and /orders
   • Check Dynamo tables and see your things
   • When an order is placed it will update the quantity of cookies left
5. Authentication
   • The orders endpoint is protected by IAM (AWS sigv4)
   • To use, go to the IAM User created, create an access key and secret, then add to the request