hpc-runtime-doctor

name	hpc-runtime-doctor
description	Diagnose HPC runtime and scheduler problems for materials simulations, including MPI/OpenMP/GPU layout, modules, CUDA/Kokkos hints, scratch paths, walltime, job arrays, restart strategy, scheduler portability, and resource mismatch. Use when an ALREADY-SUBMITTED job failed, ran slowly, got killed, or behaved differently on a cluster than on a workstation. To author or vet a NEW submission script before running, use slurm-job-script-generator instead.
allowed-tools	Read, Bash, Write, Grep, Glob
metadata	{"author":"HeshamFS","version":"1.1.3","security_tier":"high","security_reviewed":true,"tested_with":["claude-code"],"last_evaluated":"2026-06-23","eval_cases":3,"last_reviewed":"2026-06-24","standards":["SchedMD SLURM sbatch / GRES specification (--gpus, --gpus-per-node, --gres=gpu:N, --cpus-per-task)","MPI Standard (MPI Forum) process/rank model","OpenMP API specification (OMP_NUM_THREADS thread/task affinity)","NVIDIA CUDA programming model and Kokkos performance-portability backend"]}

HPC Runtime Doctor

Goal

Turn cluster symptoms into a resource-layout diagnosis, environment checklist, and safe retry plan.

Requirements

• Python 3.10+
• No external dependencies
• Works on Linux, macOS, and Windows

Inputs to Gather

Input	Description	Example
Scheduler	SLURM, PBS, LSF, local	slurm
Nodes/tasks/threads	Runtime layout	2 nodes, 128 tasks, 2 threads
GPUs	Total (whole-job) GPUs via --gpus, or per node via --gpus-per-node	--gpus 4 or --gpus-per-node 1
Symptoms	Observed failure	oom,killed,slow-gpu
MPI/OpenMP/GPU use	Parallel modes	mpi+openmp+gpu
Walltime	Requested time	12:00:00
Scratch	Whether scratch is used	true

Decision Guidance

• Check resource layout before changing physics settings.
• Confirm module/compiler/MPI/CUDA consistency before debugging solver behavior.
• Treat missing restart files and scratch cleanup as workflow failures, not physics failures.
• For GPU jobs, confirm the executable was built with the requested accelerator backend.

Script Outputs

scripts/hpc_runtime_doctor.py emits:

• resource_layout (includes tasks_per_node, total_cpus, total gpus, and gpus_per_node)
• diagnoses
• environment_checks
• retry_plan
• scheduler_notes
• warnings (layout flags such as ranks-per-GPU oversubscription, OpenMP/thread mismatch, and uneven task placement)

In default (non-JSON) mode the script also prints the resource-layout summary, any warnings, environment checks, and retry plan, so the most actionable items are never hidden.

Workflow

--gpus is the total (whole-job) GPU count. Use --gpus-per-node (SLURM --gres=gpu:N semantics) when you know the per-node allocation; total GPUs are then gpus_per_node * nodes and it overrides --gpus.

python3 skills/hpc-deployment/hpc-runtime-doctor/scripts/hpc_runtime_doctor.py \
  --scheduler slurm \
  --nodes 2 \
  --tasks 128 \
  --cpus-per-task 2 \
  --gpus 4 \
  --symptoms oom,slow-gpu \
  --uses-mpi \
  --uses-openmp \
  --uses-gpu \
  --json

The example above shares 128 ranks across 4 GPUs (32 ranks/GPU), so the warnings list surfaces Many MPI ranks per GPU (32.0 ranks/GPU) may reduce GPU efficiency. The ranks-per-GPU check uses total ranks over total GPUs, so it fires correctly on multi-node jobs (the threshold is 16 ranks/GPU).

Error Handling

Invalid resource counts stop with exit code 2. Unknown symptoms are preserved as custom items for human review.

Limitations

This skill does not query a live scheduler. It diagnoses from the submitted layout and symptoms.

Verification checklist

• Recorded the script's resource_layout block and confirmed tasks_per_node is an integer (no fractional value) and total_cpus equals tasks * cpus_per_task; if tasks_per_node is fractional, the uneven-placement warning was triaged before retrying.
• For GPU jobs, recorded the resolved total gpus (and gpus_per_node when set) and computed ranks/GPU = tasks / gpus, confirming it is at or below the 16 ranks/GPU threshold or that the resulting Many MPI ranks per GPU warning was deliberately accepted.
• Reviewed every entry in the warnings list (OpenMP-with-cpus_per_task=1, GPU-requested-but-zero-GPUs, tasks < nodes, uneven placement, scratch-for-heavy-I/O) and resolved or justified each one rather than ignoring it.
• Completed the environment_checks items as real artifacts: captured the module list, executable path/version, MPI launcher-vs-library match, accelerator build flags (CUDA/Kokkos/OpenMP), and scheduler stdout/stderr.
• Mapped each observed symptom to a diagnoses entry and verified no symptom landed in the custom category unaddressed (every custom item had stderr/stdout/module list/command line collected for human review).
• Followed the retry_plan: reran the smallest reproducing case, changed exactly one resource variable, enabled restart/checkpoint, and saved the scheduler script plus environment snapshot alongside the results.

Common pitfalls & rationalizations

Tempting shortcut	Why it's wrong / what to do
"It ran without crashing, so the layout is fine."	Run completion is not correctness. Review the warnings list and resource_layout -- oversubscription, uneven placement, or an idle GPU can silently slow or corrupt results without a crash.
"Per-node ranks fit the GPUs, so there's no oversubscription."	Oversubscription is total ranks over total GPUs, not per-node. The script computes tasks / gpus; a multi-node job can hide a high ranks/GPU value that only the unit-consistent check exposes.
"I passed --gpus, so per-node GPU count doesn't matter."	--gpus is the whole-job total. If the cluster allocates per node, use --gpus-per-node (SLURM --gres=gpu:N); it overrides --gpus and total becomes gpus_per_node * nodes. Mixing them up misreports ranks/GPU.
"The job was killed, so it's a physics/solver bug."	killed/oom/timeout are scheduler and resource categories, not physics. Check walltime, memory limits, and preemption from stdout/stderr before touching simulation parameters.
"An unknown symptom isn't in the rules, so I can skip it."	Unknown symptoms become custom diagnoses, not no-ops. Collect scheduler stderr/stdout, the module list, and the command line for human review -- silence is not a clean bill of health.
"Just change ranks, threads, and the build together to fix it faster."	Changing multiple variables at once makes the failure undiagnosable. The retry_plan mandates one variable at a time on the smallest reproducing case.

Security

Input Validation

• Inputs are scalar CLI values and booleans only; there is no free-form code path.
• Resource counts (--nodes, --tasks, --cpus-per-task, --gpus, --gpus-per-node) are validated as integers (booleans rejected), required to be non-negative and finite, and capped at 1,000,000. --nodes, --tasks, and --cpus-per-task must additionally be at least 1. Out-of-range, non-integer, or zero values exit with code 2.
• The --symptoms string is capped at 64 comma-separated entries of at most 64 characters each; --walltime is capped at 32 characters. Oversized input exits with code 2.
• Symptoms are split, trimmed, and lower-cased. Unknown symptoms are not rejected: they are preserved as custom diagnoses for human review.
• --scheduler is accepted as a free-form string and is not checked against an allowlist; it is only echoed back in the resource layout.

File Access

• The script reads and writes no files. All I/O is CLI args in and stdout out (indented JSON with --json, otherwise a human-readable summary); errors go to stderr.
• Because no paths are accepted or opened, there is no filesystem traversal surface and no path-sandboxing concern.

Tool Restrictions

• allowed-tools is Read, Bash, Write, Grep, Glob.
• Bash is used only to run the bundled scripts/hpc_runtime_doctor.py.
• Read, Grep, and Glob are used to inspect the skill's own files and any logs or submission scripts the user points at; Write is used to record diagnosis notes or a retry plan when asked.

Safety Measures

• No eval, exec, os.system, or subprocess; the script does not launch a scheduler or any external process and does not inspect environment variables.
• Argument parsing is handled by argparse, and machine-readable output is emitted as JSON.
• DoS exposure is bounded by the resource-count cap (1,000,000), the symptom caps (64 entries x 64 characters), and the walltime cap (32 characters).

References

• See references/hpc_runtime_patterns.md for scheduler and runtime diagnosis patterns.

Version History

• 1.1.3: Added a Verification checklist (evidence-based items tied to resource_layout, warnings, ranks/GPU, environment_checks, diagnoses, and the retry_plan) and a Common pitfalls & rationalizations table.
• 1.1.1: Discriminating evals -- each case now pins the script's specific output (exact ranks-per-GPU warning, diagnosis categories, resource-layout fields) via deterministic script_checks.
• 1.1.0: Unit-consistent ranks-per-GPU warning (total ranks / total GPUs), new --gpus-per-node argument, integer tasks_per_node with an uneven-placement warning, full human-readable (non-JSON) output, and input caps for resource counts, symptoms, and walltime.
• 1.0.0: Initial HPC runtime diagnosis skill.