Async Backlog Limiter

A lightweight, zero-dependency Python library to limit concurrent execution and queue size of asynchronous tasks — simulating a TCP-style backlog for async servers, job runners, or microservices to ensure overload protection.

📦 Installation

pip install async-backlog-limiter

🛠️ Usage

• If more than capacity concurrent requests are running, additional ones will wait.
• If the total number of active + waiting requests exceeds queue_limit, new ones are immediately rejected with a RateLimitExceeded exception.

import asyncio
from async_backlog_limiter import AsyncBacklogLimiter, RateLimitExceeded

limiter = AsyncBacklogLimiter(capacity=5, queue_limit=10)

async def handle_request():
    try:
        async with limiter():
            # Your async work here
            await asyncio.sleep(1)
    except RateLimitExceeded:
        print("Request rejected due to overload.")

# Run many tasks
await asyncio.gather(*[handle_request() for _ in range(20)])

# Get current statistics about the limiter's state
limiter.stats()

🧪 Tests

This project includes a comprehensive test suite using unittest.

To run tests:

python -m unittest discover -s tests