Python Serverless API

Boilerplate Flask app that is portable between different serverless platforms.

---

Platforms

Deployment and application code adaptors are being added for the following:

Platform	Deployment	Status
AWS Lambda	AWS SAM	✔️
AWS Lambda	Terraform	✔️
Azure Functions	Terraform
Google Cloud Functions	Terraform
Google Kubernetes Engine	gcloud & kubectl

Platform	Adaptor	Code/Config
Local Development	None	💾
AWS Lambda Python >= 3.6	Flask-Lambda-Python36	💾
AWS Lambda Python <= 3.6	Flask-Lambda	💾
Azure Functions
Google Cloud Functions

---

Setup

1. Create .env file and update contents

This is used to set the environment variables required for deployment and local development.

$ cp .env.example .env
$ vim .env

2. Create a virtualenv then install requirements:

$ make env
$ source env/bin/activate
$ make deps

---

Example Usage

1. Set Environment

Ensure you have created your virtualenv and have the necessary environment variables set (see setup instructions above).

$ source env/bin/activate
$ source .env

2. Run server

On host

$ make server-debug

In docker

$ docker-compose up

3. Manually test development server

$ http-prompt localhost:5000
GET /artists

---

AWS Lambda

Terraform Deployment

Ensure you have created your virtualenv and have the necessary environment variables set (see setup instructions above).

Setup

Create terraform state bucket.

$ aws s3 mb --region eu-west-2 s3://<bucket_name>

Update bucket name in /terraform/main.tf.

terraform {
  backend "s3" {
    bucket = "<bucket_name>"
    key    = "terraform.tfstate"
    region = "eu-west-2"
  }
}

Deploy

Bundle the app into a zip and deploy it using terraform.

$ ./bin/deploy

Manually Test API

$ http-prompt $(cd terraform && terraform output api_url)
GET artists

AWS Serverless Application Model (SAM) Deployment

Unlike Terraform SAM doesn't upload the zip bundle so do this using the aws-cli tool.

$ aws s3 mb s3://<mybucket>
$ aws s3 cp terraform/dist/python-serverless-api.zip s3://<mybucket>/python-serverless-api.zip

Update the S3 bucket value in the SAM config.

# template.yaml
AWSTemplateFormatVersion: '2010-09-09'
Transform: 'AWS::Serverless-2016-10-31'
Description: 'Boilerplate Python 3.6 Flask App.'
Resources:
  FlaskAPI:
    Type: 'AWS::Serverless::Function'
    Properties:
      CodeUri: s3://<mybucket>/flask-app.zip

Deploy the SAM template with Cloudformation.

$ aws cloudformation deploy \
		--template-file template.yaml \
		--stack-name python-serverless-stack-sam
		--capabilities CAPABILITY_IAM

---

Test

$ make test
$ make lint