Install NCCL Inspector and NCCL Inspector PreConf into images#2553
Open
dstaroff wants to merge 7 commits into
Open
Install NCCL Inspector and NCCL Inspector PreConf into images#2553dstaroff wants to merge 7 commits into
dstaroff wants to merge 7 commits into
Conversation
…into Slurm Check Job
dstaroff
changed the title
[slurm_]training_diag images with NCCL Inspector (+PreConf) support
asteny
reviewed
May 28, 2026
| set -e # Exit immediately if any command returns a non-zero error code | ||
|
|
||
| apt-get update | ||
| apt -y install --reinstall spank-nccl-inspector-preconf |
Collaborator
There was a problem hiding this comment.
But you have ansible role for that. Why are you using bash here?
asteny
reviewed
May 28, 2026
| rm /opt/bin/install_nccld_debug_plugin.sh | ||
|
|
||
| # Install NCCL Inspector PreConf SPANK plugin | ||
| COPY images/common/scripts/install_spank_nccl_inspector_preconf.sh /opt/bin/ |
Collaborator
There was a problem hiding this comment.
hm... but this package already in a base image and installed with ansible...
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Problem
NCCL Inspector and NCCL Inspector PreConf plugins are needed to be installed for an easy use.
plugstack.conf(matter of future PRs)Tickets to cover
Solution
This PR introduces installation of the mentioned plugins from base image root FS into the Jail under proper locations.
Along with images for Login and Worker, it provides installation of the plugin into Slurm Check Job image, and disables the NCCL Inspector PreConf plugin by default.
The change in base image versions also brings new versions of
libnccl2andlibnccl-devpackages for better performance and stability.Testing
Tested the presence of the mentioned plugins on a dev cluster.
Release Notes
Feature: Updated the version of
libnccl2andlibnccl-devpackages to improve performance and stability. Brings NCCL P2P metrics collection support for the NCCL Inspector.Feature: Added NCCL Inspector profiler plugin for deep NCCL performance investigation.
Feature: Added NCCL Inspector PreConf SPANK plugin for simple pre-configuration of NCCL Inspector for all Slurm jobs.