BentoML timeout middleware enforces 60-second default regardless of configured timeout for mounted FastAPI apps

When mounting a FastAPI application to BentoML using @bentoml.asgi_app decorator, the traffic timeout configuration (both in service decorator and --timeout CLI flag) is not applied. Requests timeout after 60 seconds with error 'Not able to process the request in 60.0 seconds' even when timeout is explicitly set to higher values like 3600 seconds.

Verify the timeout value is being read from config in the TimeoutMiddleware class at /bentoml/_internal/server/http/traffic.py. Add debug prints to check self.timeout value in the middleware. The timeout should be retrieved from the config value, not hardcoded. As a workaround, the loop.call_later(self.timeout, self._set_timer_out, waiter) line can be temporarily commented out to disable timeout enforcement, but this is not recommended for production as it disables all timeout protection.