My handler triggered an asyncio slow_callback_duration log

[wencan]

The log is as follows:

2024-08-14 22:09:18,174 asyncio WARNING base_events.py:1917 _run_once Executing <Task finished name='Task-36' coro=<RequestHandler._handle_request() done, defined at /home/wencan/Projects/codeup.aliyun.com/yswl/careai/caresrv/.venv/lib/python3.11/site-packages/aiohttp/web_protocol.py:446> result=(<Response OK eof>, False) created at /home/wencan/Projects/codeup.aliyun.com/yswl/careai/caresrv/.venv/lib/python3.11/site-packages/aiohttp/web_protocol.py:537> took 0.175 seconds

The relevant asyncio code is as follows:

        # In debug mode, if the execution of a callback or a step of a task
        # exceed this duration in seconds, the slow callback/task is logged.
        self.slow_callback_duration = 0.1

       ... ...

        dt = self.time() - t0
        if dt >= self.slow_callback_duration:
            logger.warning('Executing %s took %.3f seconds',
                           _format_handle(handle), dt)

The main logic of the handler is to generate charts using Matplotlib.
Does anyone have any good optimization suggestions?

[Dreamsorcerer]

We know the asyncio code, it's your code we need to see...

At a blind guess, you're probably doing blocking synchronous calls to matplotlib. You might be able to move the most CPU-bound calls to another thread with asyncio.to_thread(). This will allow the main code to yield back to the event loop and continue processing other tasks.