Cinema DB API

A simple REST API to store and query Movie information

You can view the hosted demo here: Live Demo

Requirements

• An AWS Account
• CircleCI Cloud (or any other CI/CD)
• Terraform 1.0.5 (Recommended)
• Node 14.7
• Mongo 5.0.2
• Docker Compose

Instructions

The following sections describe steps to configure each component of the project in order

Setting up locally

Running manually

1. Ensure you have mongoDB setup and running locally
2. Fork and git clone the project to your local machine
3. Add an .env with value for MONGO_DB_URI and PORT in the root folder
4. Run npm run start to run the application

Running with docker-compose

1. Fork and git clone the project to your local machine
2. Add an .env with value for MONGO_DB_URI in the root folder
3. Run docker-build . -t <anyname>:<anytag>
4. Update app:image to your <anyname>:<anytag> in docker-compose.yaml and run docker-compose up
5. App should now be running on http://localhost:80/

Additional commands

• npm run dev to run the app with nodemon
• npm run test to run unit tests
• npm run lint to run a code analysis

Setting up Infrastructure via Terraform

This section explains how to deploy required infrastructure via Terraform CLI locally

1. Fork and git clone the project to your local machine
2. Set the variables specified in terraform.tfvars in /terraform
3. IMPORTANT : Replace the bucket name in the modules terraform backend "s3" and resource "aws_s3_bucket" "terraform_state" of main.tf with a unique value, otherwise the following steps will fail
4. Comment out the terraform backend "s3" module and run terraform init and terraform apply (The infrastructure needs to be deployed prior to setting up the backend)
5. Uncomment the terraform backend "s3" module and run terraform init -backend-config="access_key=your aws access key" -backend-config="secret_key=your aws secret key"
6. Run terraform apply

Setting up the pipeline

This sections describes setting up the pipeline on CircleCI cloud

1. Login to CircleCI cloud and select your forked project, select main branch for config.yml in the popup
2. Set the variables specified in .env.example with your values to Project Settings > Environment Variables of the CircleCI project
3. Setup branch protection rules in settings to your liking
4. After inital deployment, you can test the application via https://{EC2-Public-IPv4-DNS-here}/movie/{some value} (Swagger will not run as SSL is not setup, check Manual Steps Required section)

Manual Steps Required

• Adding TLS/SSL needs to be manually configured, this can be achieved either by using an Elastic Load Balancer with a certificate from AWS Certificate Manager or by using a service such as LetsEncrypt or any other method of choice
• Obtaining a domain (The demo domain was obtained from freenom) and configuration (via AWS Route53 or any preferred method)

CI Process

A brief description on the CI process setup in the project

• A build will be run for every Pull Request made to the main branch, and the follwing events will occur
  • Update packages with npm ci and save-cache, run eslint for code analysis and run unit tests with mocha
  • If above steps pass, docker build will be run for additional validation
  • If all the checks are passed only, will the Pull Request merging be allowed
• A deployment will run for each Pull Request merged to master, consisting of the following steps
  • Rebuild Dockerimage and push to Docker Hub with tag set to last git commit hash (this avoid issues when updating docker-compose on server in future steps)
  • Setup AWS CLI and open port 22 via Security Group for connecting to EC2 instance
  • Run scp and copy docker-compose.yml
  • Update running docker-compose instance via ssh along with updated image tag, the deployment completes
  • Close the port 22 previously opened to connect to the EC2 instance

Notes

• Should a step between Opening and Closing Port 22 for SSH & SCP fail , the EC2 instance will be left exposed to the IP used by CircleCI pipeline, 'when: always' was added to the close port step as a potential fix for this
• Some of the package and file imports may look inconsistent in the source due to the use of in-built experimental ES6 features packaged in node by adding {"type":"module"} in package.json without setting up a transpiler such as babel