Deployment of Flask application on Amazon Fargate using Terraform

The project describe how to deploy a Flask application to Amazon Fargate using Terraform. The application is using AWS Cognito for user authentication. It is also able to send a message to AWS SQS when visiting specific URL. The message triggers an AWS lambda function.

The purpose of the project was to:

• learn how to deploy a Flask application to AWS
• familiarize myself with some AWS services and Python boto3 library
• learn about Terraform tool

Terraform

Terraform files in the project are able to deploy and configure following AWS services:

• VPC - create VPC network
• IAM - create roles and policies
• ECS - deploy Fargate service
• Congito - user authentication
• SQS - create a queue
• CloudWatch - logs monitoring
• S3 - store Lambda function
• Lambda - function will be triggered by new message in SQS

Setup

• Clone repository
• Create AWS configuration and credential file: https://docs.aws.amazon.com/cli/latest/userguide/cli-configure-files.html
• Change aws_region variable in terraform/0-variables.tf according to your settings
• Initialize a Terraform working directory

cd terraform
terraform init

• Deploy application

terraform apply

• Destroy application

terraform destroy

NOTE

terraform/0-variables.tf and flask_app/config.ini contains some common settings. Modification of one file requires the same change to be made to the other file.

Usage

Open AWS ECS service and grab public IP address of deployed container. Application is running on port 5008

Routes

Main page

/

Login page. Credentials for Cognito user are defined in flask_app/config.ini

/login

Protected page. It can be accessed only by logged in users

/protected

Visiting this route will send a message to SQS. Then the message will trigger Lambda function

/queue

Logs

Logs are stored in AWS CloudWatch:

• application logs /ecs/flask_app
• Lamba function logs /aws/lambda/triggered_by_queue

Technologies

• Python 3.8.0
• Terraform 0.12.21
• Flask 1.1.1
• boto3 1.12.17
• AWS
• Docker