Atelier Products API

About

The goal of this project was to build a scalable RESTful API service for the Products component of an e-commerce website. The service was optimized by leveraging database indexing, connection pooling, and Redis caching. The server and Postgres database were containerized with Docker and scaled using Docker Swarm.

Development

1. Clone this repo https://github.com/sdc-cirrus/products-api.git
2. Install Docker
3. Create a .env file in the project's root directory and provide POSTGRES_USER, POSTGRES_PASSWORD, POSTGRES_DB, POSTGRES_HOST and POSTGRES_PORT.
4. In the project's root directory, run:

docker-compose up

Tech Stack

• Docker
• PostgreSQL
• Redis
• Node-PG
• Artillery
• LoaderIO

ETL

1. Inherited legacy dataset consisting of 40M records.
2. Used Node streams to transform data.
3. Designed appropriate schemas for the data
4. Leveraged Node-PG connection pools and DBeaver to load the data into the local development database and deployed database.
5. Incorporated Docker Swarm to replicate multiple instances of the server and scale the API

Testing

Used Loader.IO and Artillery to stress test the API.

Routes

HTTP METHOD	Endpoint	Returns	Status
GET	/products/:productId/related	Returns the IDs of products related to the specified product	200
GET	/products/:productId/styles	Returns the all styles available for the given product	200
GET	/products/:productId	Returns all product level information for a specified product id	200
GET	/products/list	Retrieves the list of products	200