Async Rate Limiter (Sliding Window + Lua + Redis)

A Python async rate limiter using Redis and Lua with sliding window algorithm.
Supports retries, exponential backoff, and optional exception handling.

Installation

pip install redis asyncio

Usage

Basic usage with inline function

import asyncio
import redis.asyncio as aioredis
from rate_limit_module import RateLimit  # replace with your module
from rate_limiter.exceptions import RetryLimitReached

redis_client = aioredis.Redis(host='localhost', port=6379, db=0)

rate_limit = RateLimit(
    redis=redis_client,
    limit=5,
    window=10,
    retries=3,
    backoff_ms=200,
    backoff_factor=2.0,
    retry_on_exceptions=(ValueError,),
)

async def my_task():
    print('Task executed.')
    return 42

wrapped = rate_limit(fn=my_task, key='task_key')

try:
    result = await wrapped()
    print(result)
except RetryLimitReached:
    print('All retry attempts exhausted.')

Using as a decorator

rate_limit = RateLimit(
    redis=redis_client,
    limit=3,
    window=5,
    retries=4,
    backoff_ms=100,
    backoff_factor=1.5,
)

@rate_limit(key='decorated_task')
async def my_decorated_task():
    print('Decorated task executed.')
    return 'done'

try:
    await my_decorated_task()
except RetryLimitReached:
    print('Rate limit retry attempts exhausted.')

Exception behavior

After all retries are used without success, the limiter raises RetryLimitReached exception
instead—making it easier to handle failure explicitly in your code.

Features

• Sliding window rate limiting using Redis + Lua
• Async-friendly
• Retries with exponential backoff configurable
• Optional exception-based retry logic
• Raises RetryLimitReached when retry attempts are exhausted
• Supports both inline wrapper and decorator syntax

License

MIT License