[dev.]
←Back to selected work
Case Study · backend

RateGuard – Distributed Rate Limiting Service

Distributed rate limiting microservice architecture enforcing multi-tier limits at sub-5ms latency with atomic Redis Lua scripts.

Track

backend

Technologies

Spring Boot, Redis Lua, PostgreSQL, Docker, React

Primary Impact

Token Bucket & Sliding Window algorithms

RepositoryView on GitHub ↗

Context & Challenge

High-throughput microservices require resilient, low-latency distributed rate limiters to protect upstream APIs from abuse without introducing latency bottlenecks.

Process & Implementation

Implemented Token Bucket and Sliding Window algorithms with atomic Redis Lua scripts across 3 instances, enforcing 3-tier limits (100/1K/10K req/min); built a React dashboard with live WebSocket traffic graphs and PostgreSQL counter recovery on Redis restarts.

Verified Outcomes & Key Facts

  • Token Bucket & Sliding Window algorithms
  • sub-5ms latency across 3 instances
  • 0 race conditions over 10K+ load tests
  • 0 limit breaches on 500+ concurrent connections