PyPI 官网下载 | aio-throttle-1.0.0.tar.gz

This library is inspired by [this book](https://landing.google.com/sre/sre-book/chapters/handling-overload/). Features: 1. Set capacity(max parallel requests) and queue(max queued requests) limits. 1. Per-consumer limits. For instance, to not allow any consumer to use more than 70% of service's capacity. 1. Per-request priorities. For instance, to not allow requests with lowest priority to be queued or to now allow requests with normal priority to use more than 90% of service's capacity. Example: ```python from aio_throttle import Throttler, MaxFractionCapacityQuota, ThrottlePriority, ThrottleResult capacity_limit = 100 queue_limit = 200 consumer_quotas = [MaxFractionCapacityQuota(0.7)] priority_quotas = [MaxFractionCapacityQuota(0.9, ThrottlePriority.NORMAL)] throttler = Throttler(capacity_limit, queue_limit, consumer_quotas, priority_quotas) async with throttler.throttle(consumer, priority) as result: ... # check if result is ThrottleResult.ACCEPTED or not ``` Example of integration with `aiohttp`: ```python from typing import Awaitable, Callable from aiohttp import web from aiohttp.web_middlewares import middleware from aiohttp.web_request import Request from aiohttp.web_response import StreamResponse from aio_throttle import Throttler, ThrottlePriority Handler = Callable[[Request], Awaitable[StreamResponse]] @middleware async def middleware(request: Request, handler: Handler) -> StreamResponse: throttler: Throttler = request.app['throttler'] service = request.headers.get('X-Service', 'UNKNOWN') priority = ThrottlePriority(request.headers.get('X-Request-Priority', ThrottlePriority.NORMAL)) async with throttler.throttle(service, priority) as result: if not result: return web.Response(status=503) return await handler(request) ```