Reviews System Architecture

The goal of this project was to take a small dataset and scale it to 10 million documents on a horizontally scaled AWS EC2 Deployment. Using Two Linux EC2 Microservers with elastic IPs, NGINX Round Robin Loadbalancing, as well as NGINX's built-in caching, I was able to achieve 6,000 Requests per Second, 99.75% success rate, and a 76ms response time.

These results were tested locally using Artillery.io and then online through New Relic

Screens

NewRelic Cloud results Artillery local performance Dataset of Reviews CSV - can provide file on request.

Usage

• Fork and clone repo(s)

Install dependencies

• npm install

Start up microservice

1. npm run start-main (seed the database and starts the server)
2. npm run build

To view, open browser and visit localhost: 3001

Deployed Site no longer up, but feel free to reach out if you'd like to test. Deployed site: http://ec2-18-191-56-236.us-east-2.compute.amazonaws.com:3001/