Skip to content

feat: report per-node validation detail - #1378

Merged
michael-balint merged 1 commit into
masterfrom
dholt/add-validator-per-node-detail
Aug 17, 2026
Merged

feat: report per-node validation detail#1378
michael-balint merged 1 commit into
masterfrom
dholt/add-validator-per-node-detail

Conversation

@dholt

@dholt dholt commented Aug 14, 2026

Copy link
Copy Markdown
Contributor

Add per-node detail to DeepOps validators

What this is

Implement the 26.09 public validator improvement: add stable per-node detail to the --json output of validate_slurm.py and validate_k8s.py while preserving existing aggregate fields and exit behavior. Slurm detail must report each unique node's normalized state and configured GRES GPU count. Kubernetes detail must report each node's Ready status and allocatable nvidia.com/gpu count, handling malformed values as zero. Add focused stdlib unit tests, update the validator skill only where needed, and keep human output compatible. Do not access a live scheduler or cluster; use parser fixtures only. Do not push or open a PR.

Verification

  • Deterministic checks passed: diff check, validation tests, python compile, public sanitizer, parent commit
  • Two independent agent reviews approved head abf0007eb062 (different model vendors; strict verdict contract)
  • Published by the DeepOps maintenance loop's deterministic publisher after its publication gate (reviews, provenance, exact-head) passed

@dholt
dholt requested a review from michael-balint August 14, 2026 19:21
@michael-balint
michael-balint merged commit 6130d9f into master Aug 17, 2026
30 checks passed
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

2 participants