Enable NUMA-aware scheduling for H100 4-GPU runner (#696)#740
Draft
georgehong wants to merge 5 commits into
Draft
Enable NUMA-aware scheduling for H100 4-GPU runner (#696)#740georgehong wants to merge 5 commits into
georgehong wants to merge 5 commits into
Conversation
This was referenced Jun 11, 2026
tofu plan — arc-cbr-production✅ Plan succeeded · commit Plan output |
tofu plan — arc-cbr-production-uw1✅ Plan succeeded · commit Plan output |
tofu plan — meta-prod-aws-ue1✅ Plan succeeded · commit Plan output |
georgehong
added a commit
that referenced
this pull request
Jun 11, 2026
Enables nfd + numa-scheduler on both prod clusters in clusters.yaml and wires scheduler_name: numa-scheduler into the H100 4-GPU runner definition (l-x86iamx-88-900-h100-4). Per-runner-def scheduler_name in generate_runners.py allows targeting specific runners without affecting others. Workflow pods for this runner use the numa-scheduler, which checks NRT data to ensure a single NUMA zone can satisfy the full 4-GPU request — preventing TopologyAffinityError on fragmented p5.48xlarge nodes. Other runners are unaffected (no scheduler_name in their defs = default scheduler). 8-GPU H100 jobs on p5-large use best-effort topology and must NOT use the numa-scheduler. Part of #696 ghstack-source-id: 3916a52 Pull-Request: #740
georgehong
added a commit
that referenced
this pull request
Jun 11, 2026
Enables nfd + numa-scheduler on both prod clusters in clusters.yaml and wires scheduler_name: numa-scheduler into the H100 4-GPU runner definition (l-x86iamx-88-900-h100-4). Per-runner-def scheduler_name in generate_runners.py allows targeting specific runners without affecting others. Workflow pods for this runner use the numa-scheduler, which checks NRT data to ensure a single NUMA zone can satisfy the full 4-GPU request — preventing TopologyAffinityError on fragmented p5.48xlarge nodes. Other runners are unaffected (no scheduler_name in their defs = default scheduler). 8-GPU H100 jobs on p5-large use best-effort topology and must NOT use the numa-scheduler. Part of #696 ghstack-source-id: cf8614e Pull-Request: #740
georgehong
added a commit
that referenced
this pull request
Jun 11, 2026
Enables nfd + numa-scheduler on both prod clusters in clusters.yaml and wires scheduler_name: numa-scheduler into the H100 4-GPU runner definition (l-x86iamx-88-900-h100-4). Per-runner-def scheduler_name in generate_runners.py allows targeting specific runners without affecting others. Workflow pods for this runner use the numa-scheduler, which checks NRT data to ensure a single NUMA zone can satisfy the full 4-GPU request — preventing TopologyAffinityError on fragmented p5.48xlarge nodes. Other runners are unaffected (no scheduler_name in their defs = default scheduler). 8-GPU H100 jobs on p5-large use best-effort topology and must NOT use the numa-scheduler. Part of #696 ghstack-source-id: 7192c6c Pull-Request: #740
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
1 participant
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):
Enables nfd + numa-scheduler on both prod clusters in clusters.yaml
and wires scheduler_name: numa-scheduler into the H100 4-GPU runner
definition (l-x86iamx-88-900-h100-4).
Per-runner-def scheduler_name in generate_runners.py allows targeting
specific runners without affecting others. Workflow pods for this
runner use the numa-scheduler, which checks NRT data to ensure a
single NUMA zone can satisfy the full 4-GPU request — preventing
TopologyAffinityError on fragmented p5.48xlarge nodes.
Other runners are unaffected (no scheduler_name in their defs =
default scheduler). 8-GPU H100 jobs on p5-large use best-effort
topology and must NOT use the numa-scheduler.
Part of #696