Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Probing AM not possible if local endpoint has closed #308

Open
pentschev opened this issue Oct 23, 2024 · 0 comments
Open

Probing AM not possible if local endpoint has closed #308

pentschev opened this issue Oct 23, 2024 · 0 comments
Labels
bug Something isn't working

Comments

@pentschev
Copy link
Member

Probing AM is not possible if the local endpoint has closed, as opposed to the remote endpoint closing. This is due to ucxx::Endpoint::_handle being swapped with nullptr after ucxx::Endpoint::close{,Blocking} is called, after which is impossible to obtain the original handle to probe for AM in the worker. Additionally, the Python async interface relies on the Python core bindings to execute UCXEndpoint.am_probe() but calling Endpoint.close()/Endpoint.abort() assigns the underlying self._ep = None, after that time there are no more ways to probe for AM for that endpoint even if the handle is known since ucxx::Worker::amProbe() is not even exposed directly to the Python core bindings, but only via UCXEndpoint.am_probe().

Once this is resolved, test_message_probe should be split into two: test_message_probe_local_close and test_message_probe_remote_close (exactly what test_message_probe is today).

@pentschev pentschev added the bug Something isn't working label Oct 23, 2024
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
bug Something isn't working
Projects
None yet
Development

No branches or pull requests

1 participant