Request timing out in 60 seconds regardless of timeout-keep-alive value

Hello, I’m running uvicorn server with Quart, and when I try to upload a large file or one through a slow shared folder connection, the request always times out at 60 seconds.

I’m not running any other server like NGINX, the response is coming from Uvicorn itself

This is how I’m running it :

uvicorn.run('app:application', host="0.0.0.0", port=500, loop="asyncio", reload=True, workers=2, timeout_keep_alive=3600, log_level='debug')

It returns me 408 Request Timeout after 60s in the server logs.

Funny thing is the browser is not notified immediately, only after it finished “uploading” the whole file he’s aware of the timeout, not sure that’s just normal

Using uvicorn 0.11.5

[euri10]

reading quart docs, it seems to me the issue likely lies here: https://github.com/pgjones/quart/blob/89888426a980689c148603db21bbb3017674a4ea/docs/discussion/dos_mitigations.rst#slow-request-body

if you can confirm ?

[euri10]

you cannot send a reload signal, reload is automatically triggered once a watched file is modified, it’s more a development mode feature.

Yes I understand it’s for development

I just wanted to know if there’s a documented way of safely restarting the server in production, or do I just use a supervisor and restart the process?

Not that I know, I think most people use gunicorn for that