🦬 BuffaloMQ

A lightweight, production-grade Redis job queue for Python with a built-in real-time Web Dashboard.

BuffaloMQ provides a simple, reliable, and dependency-light alternative to heavy task queues like Celery. It leverages Redis and atomic Lua scripts for crash resilience, delayed scheduling, visibility timeouts, and DLQ management — complete with a zero-config web dashboard CLI.

🛠 Tech Stack

• Core & Broker: Python 3.9+ & Redis (via atomic Lua scripts)
• Dashboard Server: FastAPI & Uvicorn
• Dashboard Frontend: React 18, TypeScript, Vite

✨ Features

• ⚡ Lightweight & Fast: Pure Python API powered by atomic Redis Lua operations.
• 🖥 Zero-Config Web Dashboard: Launch a real-time monitoring UI with a single CLI command (buffalomq dashboard).
• 🔄 Job Retries & Backoff: Configurable retry limits and exponential backoff.
• ⏱ Delayed Jobs: Schedule tasks to run after a specific delay.
• 🔒 Crash Recovery: Visibility timeouts & heartbeat threads prevent lost jobs if a worker crashes.
• ⏸ Queue Controls & DLQ Re-drive: Pause/resume queues and re-drive failed jobs straight from the dashboard or API.

📦 Installation

pip install buffalomq

🚀 Quickstart

1. Submit Jobs (Producer)

import redis
from buffalomq import Queue

r = redis.Redis(host='localhost', port=6379, db=0)
queue = Queue("email-notifications", r)

# Enqueue immediate job
queue.add({"to": "user@example.com", "subject": "Welcome!"})

# Enqueue delayed job (runs after 30s) with 3 max retries
queue.add({"to": "user@example.com", "subject": "Reminder"}, options={"delay": 30, "max_attempts": 3})

2. Process Jobs (Worker)

import redis
from buffalomq import Worker

def process_email(job_data):
    print(f"Sending email to: {job_data['to']}")

r = redis.Redis(host='localhost', port=6379, db=0)
worker = Worker("email-notifications", r, process_email)

# Start listening for jobs
worker.run()

3. Launch Web Dashboard

buffalomq dashboard --redis redis://localhost:6379/0

Automatically opens http://127.0.0.1:8000 in your browser to monitor queues, inspect payloads, pause workers, and re-drive failed jobs.