[c10d] differentiate timeout errors from nccl errors #138240
Closed
Conversation
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Summary: It's important for c10d to differentiate different reasons of watchdog failures. E.g, timeout vs nccl errors, and let users to handle the errors depends on the type of error Test Plan: UT Subscribers: Tasks: Tags: [ghstack-poisoned]
🔗 Helpful Links🧪 See artifacts and rendered test results at hud.pytorch.org/pr/138240
Note: Links to docs will display an error until the docs builds have been completed. ✅ No FailuresAs of commit 50ac546 with merge base 20af56d ( This comment was automatically generated by Dr. CI and updates every 15 minutes. |
pytorch-bot
bot
added
oncall: distributed
Skylion007
approved these changes
Oct 17, 2024
Collaborator
Skylion007
left a comment
Skylion007
left a comment
There was a problem hiding this comment.
Thank you! This functionality will be really helpful
Contributor
Author
|
@pytorchbot merge |
Collaborator
Merge startedYour change will be merged once all checks pass (ETA 0-4 Hours). Learn more about merging in the wiki. Questions? Feedback? Please reach out to the PyTorch DevX Team |
cyyever
reviewed
Oct 18, 2024
| } catch (...) { | ||
| LOG(ERROR) | ||
| << logPrefix() | ||
| << "Failed to rerieve TORCH_NCCL_DESYNC_DEBUG report with unknown error." |
Collaborator
There was a problem hiding this comment.
Can the exception what() be logged here?
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
4 participants
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Stack from ghstack (oldest at bottom):
Summary:
Our watchdog does not differentiate timeout from NCCL errors clearly in terms of both log and code paths.
It's important for c10d to differentiate different reasons of watchdog
failures. E.g, timeout vs nccl errors, and possibly let users to handle the
errors differently depends on the type of errors
Test Plan:
UT
Subscribers:
Tasks:
Tags:
cc @XilunWu @H-Huang @awgu @kwen2501 @wanchaol @fegin @fduwjj @wz337 @wconstab @d4l3k @c-p-i-o