skalify / qa-system-design

Skalify's system design for backend operations

Questions & Answers API

General Info

• This project serves as the backend design for Project Atelier, an e-commerce company.
• This server handles the Questions & Answers of the individual products.
• An ETL process was used to parse millions of rows of data and optimize the schema.
• The database was created with PostgreSQL.
• Load balancing and Caching was implemented utilizing Nginx.

Perfomance

Back End Architecture utilizes AWS to deploy a load balancer with a cache, 4 servers, and a Postgres database. All load tests performed via Loader.io.

Typical Load Performance

Perfomance at typical load of 1000 clients per second. 1ms latency and 0% error rate.

Typical 1000 RPS Load

Optimization

Cache Implementation

Cache and load balancer implemented using Nginx - Latency increased to 7ms, but was able to push performance and be able to handle 10,000 rps. Through a cached implementation and proper load balancing, the server was able to maintain a 0% error rate.

Performance Metrics

Tech Stack

Questions & Answers: Mo Akbari