This repository explores the implementation and performance of rate-limiting algorithms, focusing on their impact when used in a reverse proxy setting. The study was conducted as part of CS 8395 - Advanced Topics in Software Engineering, with an emphasis on real-world scenarios involving varying system loads.

• Algorithms Explored:
  • Token Bucket
  • Leaky Bucket
  • Fixed Window
  • Sliding Window
• Core Features:
  • Reverse Proxy Functionality:
    • The rate-limiting algorithms were implemented to regulate traffic as a reverse proxy, ensuring controlled throughput while handling multiple client requests.
  • Load Conditions:
    • Evaluated performance under burst loads and constant loads, simulating real-world scenarios.
  • Metrics Monitored:
    • Power consumption.
    • Throughput under stress and normal operation.

• Tech Stack: Written in Go and deployed on Raspberry Pi devices for distributed testing.
• Command-Line Configuration:
  • Allows selection of rate-limiting algorithms and tuning of parameters such as token replenishment rates and burst capacities.
• Distributed Setup:
  • Server and client run on separate Raspberry Pi devices.
  • Algorithms run as independent services to simulate modular deployment.

The final project report, detailing implementation, experiments, and findings, is available upon request. Please contact us if you are interested.