You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
I've seen this quite a few times now in unit tests, for example in aiidateam/aiida-core#4775 (comment)
This coroutine is created in plumpy.futures.create_task, which is called by plumpy.communicator.convert_to_comm, that is used when adding an rpc/task/broadcast subscriber in plumpy.communicator.LoopCommunicator.
It is weird that it should say it was not awaited, because it is wrapped in asyncio.run_coroutine_threadsafe(run_task(), loop), so you would think that it should always be awaited.
Is this just because the loop is closed while run_task() is running (I guess these tasks should run forever until the subscriber is removed), or does it point to a deeper problem?
Could LoopCommunicator.close ensure such tasks are stopped gracefully, e.g. by ensuring all subscribers are removed.
I also curious why this happened and sorry that I have no opinion on how to
fix this. 😭 I remember this message first time appeared was after I put
the run_task in the thread safe.
As for your guess, when and why the loop here is being closed? now the
Runner has only one loop in its lifetime.
On Fri, Feb 26, 2021 at 17:25 Chris Sewell ***@***.***> wrote:
I've seen this quite a few times now, for example in aiidateam/aiida-core#4775
(comment)
<aiidateam/aiida-core#4775 (comment)>
This coroutine is created in plumpy.futures.create_task, which is called
by plumpy.communicator.convert_to_comm, that is used when adding an
rpc/task/broadcast subscriber in plumpy.communicator.LoopCommunicator.
It is weird that it should say it was not awaited, because it is wrapped
in asyncio.run_coroutine_threadsafe(run_task(), loop), so you would think
that it should always be awaited.
Is this just because the loop is closed while run_task() is running (I
guess these tasks should run forever until the subscriber is removed), or
does it point to a deeper problem?
Could LoopCommunicator.close ensure such tasks are stopped gracefully,
e.g. by ensuring all subscribers are removed.
cc @unkcpz <https://github.com/unkcpz>
—
You are receiving this because you were mentioned.
Reply to this email directly, view it on GitHub
<#211>, or unsubscribe
<https://github.com/notifications/unsubscribe-auth/ACDMFRPC2EO2NBJHCXHRU5TTA5SHHANCNFSM4YIDCFKA>
.
I've seen this quite a few times now in unit tests, for example in aiidateam/aiida-core#4775 (comment)
This coroutine is created in
plumpy.futures.create_task
, which is called byplumpy.communicator.convert_to_comm
, that is used when adding an rpc/task/broadcast subscriber inplumpy.communicator.LoopCommunicator
.It is weird that it should say it was not awaited, because it is wrapped in
asyncio.run_coroutine_threadsafe(run_task(), loop)
, so you would think that it should always be awaited.Is this just because the loop is closed while
run_task()
is running (I guess these tasks should run forever until the subscriber is removed), or does it point to a deeper problem?Could
LoopCommunicator.close
ensure such tasks are stopped gracefully, e.g. by ensuring all subscribers are removed.cc @unkcpz
The text was updated successfully, but these errors were encountered: