[c10d][simple] increase the default heartbeat timeout to be larger #128751
Conversation
This file contains 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: In multiple cases, we were seeing monitor heartbeat timeout triggred by commAbort, which is undesriable as it hides the real timeout reason (e.g.,, commAbort) and mislead debugging to other types of hangs (e.g., cuda calls, etc) Test Plan: CI Tags: [ghstack-poisoned]
🔗 Helpful Links🧪 See artifacts and rendered test results at hud.pytorch.org/pr/128751
Note: Links to docs will display an error until the docs builds have been completed. ⏳ No Failures, 1 PendingAs of commit 5d331c5 with merge base 0344f95 ( This comment was automatically generated by Dr. CI and updates every 15 minutes. |
pytorch-bot
bot
added
oncall: distributed
shuqiangzhang
added a commit
that referenced
this pull request
Jun 14, 2024
Summary: In multiple cases, we were seeing monitor heartbeat timeout triggred by commAbort, which is undesriable as it hides the real timeout reason (e.g.,, commAbort) and mislead debugging to other types of hangs (e.g., cuda calls, etc) Test Plan: CI Tags: ghstack-source-id: ba2bd5e7b92cbcc51672dc3c932057865b0a5a73 Pull Request resolved: #128751
fduwjj
reviewed
Jun 17, 2024
| heartbeatTimeoutInSec_ = | ||
| getCvarInt(TORCH_NCCL_HEARTBEAT_TIMEOUT_SEC, 60 * 10 /*10 Mins*/); | ||
| getCvarInt(TORCH_NCCL_HEARTBEAT_TIMEOUT_SEC, 60 * 11 /*11 Mins*/); |
There was a problem hiding this comment.
I am ok with changes, but just to be safe. Is there any workloads that we can verify that this indeed resolves the race condition? I mean why not 12 mins or longer?
|
Similar question to @fduwjj, it seems pretty unlikely that a 10% increase is a magic bullet unless we have some other 10 min timeout in the system that we are racing against. |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
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:
In multiple cases, we were seeing monitor heartbeat timeout triggred by
commAbort, which is undesriable as it hides the real timeout reason
(e.g.,, commAbort) and mislead debugging to other types of hangs (e.g.,
cuda calls, etc)
Test Plan:
CI
Tags:
cc @mrshenli @pritamdamania87 @zhaojuanmao @satgera @gqchen @aazzolini @osalpekar @jiayisuse @H-Huang @kwen2501 @awgu @penguinwu @fegin @XilunWu @wanchaol @fduwjj @wz337 @tianyu-l @wconstab @yf225 @chauhang @d4l3k