Asyncio tasks getting cancelled

[Ian Kelly]

On Mon, Nov 5, 2018 at 6:20 PM <philip.m at tura-home.de> wrote:
> Again sorry for the confusion, but I don't think this is an issue with
> restarting loops, as this isn't happening in my application.

Sure, I wasn't talking about restarting loops so much as strategies
for making sure that everything has completed in the first place.

> For context:
> https://github.com/ldo/dbussy/issues/13
> https://gist.github.com/tu500/3232fe03bd1d85b1529c558f920b8e43
>
> It really feels like asyncio is loosing strong references to scheduled
> tasks, as excplicitly keeping them around helps. Also, the error
> messages I'm getting are the ones from here:
> https://github.com/python/cpython/blob/16c8a53490a22bd4fcde2efaf4694dd06ded882b/Lib/asyncio/tasks.py#L145
> Which indicates that the tasks actually weren't even started at all?

No, it indicates that it was cleaned up (likely because the program
exited) before the task completed. Which likely implies that the loop
exited without waiting for it. From the stack trace, you're using
loop.run_until_complete(main()). Like asyncio.run, that only runs
until the specific thing you pass it completes, which might wait on
other things or might not. Anything else that's still pending is going
to be left up in the air unless you subsequently restart the same
event loop.

> Though not really related with my actual problem, so getting off topic,
> but I can imagine an architecture where that would be "There aren't any
> running tasks any more." or even "Never."

Well, "never" is easy; that's just loop.run_forever(). "No more
running tasks" is trickier; I'm not sure how one might do that without
collecting all the tasks (at least, all the top-level tasks) oneself.
Implementing a custom task factory might help.