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

Hashed trace user events should do an all-reduce at the end instead of sending pt2pts to rank 0 during the run #2387

Open
lifflander opened this issue Jan 8, 2025 · 0 comments · May be fixed by #2395
Assignees

Comments

@lifflander
Copy link
Collaborator

Describe the bug
When using trace::TraceScopedEventHash scope("XXX");, the node that calls this immediately sends a message to rank 0 with the hashed event, causing tons of messages that interfere with the application run. We should all-reduce these at the end of the run when we need to write out the STS file.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Projects
None yet
Development

Successfully merging a pull request may close this issue.

2 participants