Skip to content

Increase poll log interval to reduce log spam #541

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.

Sign up for GitHub

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

Draft
kmontemayor2-sc wants to merge 5 commits into
base: main
Choose a base branch
from
Draft

Increase poll log interval to reduce log spam #541

Changes from 2 commits
Commits
Show all changes
5 commits
Select commit Hold shift + click to select a range
390a2bc
Bump wait time to reduce log spam
kmonte Mar 11, 2026
7f5fc9c
update
kmonte Mar 11, 2026
2c678da
Merge branch 'main' into kmonte/bump-gs-polling-timeout
kmonte Mar 11, 2026
11f8f7d
update
kmonte Mar 11, 2026
d596855
update
kmonte Mar 11, 2026
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
3 changes: 2 additions & 1 deletion gigl/distributed/utils/networking.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -233,7 +233,7 @@ def wait_for_readiness_signal(
readiness_uri: Uri,
timeout: float = 3600.0,
poll_interval: float = 10.0,
Comment thread
kmontemayor2-sc marked this conversation as resolved.
Outdated
log_every_n_attempts: int = 10,
log_every_n_attempts: int = 60,
) -> None:
"""Poll for a readiness sentinel file before initiating RPC connections.

Expand All @@ -244,6 +244,7 @@ def wait_for_readiness_signal(
Supports both GcsUri (production) and LocalUri (testing).
timeout: Maximum time in seconds to wait for the signal. Defaults to 3600.
poll_interval: Time in seconds between poll attempts. Defaults to 10.
log_every_n_attempts: Number of attempts between log messages. Defaults to 60.
Copy link
Copy Markdown
Collaborator

@svij-sc svij-sc Mar 11, 2026

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Suggested change
log_every_n_attempts: Number of attempts between log messages. Defaults to 60.
log_every_n_attempts: Number of attempts between log messages. Defaults to 60. i.e. with poll_interval set to 10, and log_every_n_attempts set to 60, we will log ever 600 seconds which is every 10 minutes.

10 minutes isnt too much? Usually I consider something hanging if no logs for more than 2 -4 mins.

Copy link
Copy Markdown
Collaborator Author

@kmontemayor2-sc kmontemayor2-sc Mar 11, 2026

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

We log updates every 10 minutes, I think every compute rank (e.g. num gpus) dumping every minute is probably too frequent and kind of clogs up the logs.

I guess we can make it 5 minutes? And update the logs here to expect the next update.


Raises:
TimeoutError: If the readiness signal is not found within the timeout.
Expand Down