Set explicit 30-minute timeout on all init_process_group calls#1069
Merged
Conversation
The Gloo+torchrun and srun init paths had no explicit timeout, relying on PyTorch defaults. This makes all paths consistent with the NCCL+torchrun path, ensuring collective operations time out after 60 minutes when a peer dies instead of potentially hanging longer. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
jpdunc23
approved these changes
Apr 21, 2026
Member
jpdunc23
left a comment
jpdunc23
left a comment
There was a problem hiding this comment.
LGTM. Per discussion on Slack, maybe update to 30 min in this PR?
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
mcgibbon
changed the title
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
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
2 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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
The Gloo+torchrun and srun
init_process_grouppaths had no explicit timeout, relying on PyTorch defaults (30 minutes). This makes all paths consistent with the NCCL+torchrun path, ensuring collective operations time out after 60 minutes when a peer dies.Changes:
fme.core.distributed.TorchDistributed.__init__: addedtimeout=timedelta(minutes=60)to the Gloo+torchrun and sruninit_process_groupcallsTests added