asyncio.exceptions.CancelledError on scheduler.shutdown()

Things to check first

• I have checked that my issue does not already have a solution in the FAQ

• I have searched the existing issues and didn’t find my bug already reported there

• I have checked that my bug is still present in the latest release

Version

3.9.1

What happened?

Hi again,

So I downgraded to APScheduler v3.9.1 (from the latest alpha2) and set up a job using AsyncIOScheduler() in the Starlite API framework. This was the easiest path forward for me (RE #673)

I am able to successfully start and stop the scheduler with the Starlite on_startup and on_shutdown events. But on shutdown, I am getting exceptions thrown after the shutdown event.

Is this something I need to worry about, or is there anything I can do to handle the errors? The tasks query an API and write to a DB.

STACKTRACE:

INFO:     Shutting down
INFO:     Waiting for application shutdown.
INFO - 2022-10-27 16:13:33,785 - main - main - 🛑 Stopping task scheduler
INFO - 2022-10-27 16:13:33,786 - apscheduler.scheduler - base - Scheduler has been shut down
ERROR - 2022-10-27 16:13:33,791 - apscheduler.executors.default - base_py3 - Job "main (trigger: interval[0:01:00], next run at: 2022-10-27 16:14:08 EAT)" raised an exception
Traceback (most recent call last):
  File "/miniforge3/envs/py311/lib/python3.11/site-packages/apscheduler/executors/base_py3.py", line 30, in run_coroutine_job
    retval = await job.func(*job.args, **job.kwargs)
             ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/Documents/GitHub/data-wrangler/src/job.py", line 332, in main
    async with rate_limiter:
  File "/miniforge3/envs/py311/lib/python3.11/site-packages/aiolimiter/leakybucket.py", line 110, in __aenter__
    await self.acquire()
  File "/miniforge3/envs/py311/lib/python3.11/site-packages/aiolimiter/leakybucket.py", line 97, in acquire
    await wait_for(
  File "/miniforge3/envs/py311/lib/python3.11/asyncio/tasks.py", line 466, in wait_for
    await waiter
asyncio.exceptions.CancelledError
INFO:     Application shutdown complete.
INFO:     Finished server process [78813]
INFO:     Stopping reloader process [78810]

How can we reproduce the bug?

Here are the functions I’m using to schedule the jobs. I tried added scheduler.remove_all_jobs() before shutdown, but that created even more exceptions.

async def start_scheduler(state: State):
    scheduler = AsyncIOScheduler()
    scheduler.add_job(
        main, "interval", seconds=60, max_instances=1, id="data_job_1"
    )
    scheduler.start()
    state.scheduler = scheduler
    logger.info("🚦 Starting the task scheduler")


async def stop_scheduler(state: State):
    scheduler = state.scheduler
    scheduler.shutdown(wait=True)
    logger.info("🛑 Stopping task scheduler")

[agronholm]

Asyncio support was added to the 3.x series as an afterthought. It never worked really well, so in v4.0 I redesigned the library as async-first. I should mention that, among other things, wait=True does not work with the asyncio scheduler, as there is no way to wait for the shutdown in a synchronous function.

[agronholm]

I was hoping to see what you tried to do with AsyncExitStack, based on my advice.