slurmwatch

Live per-process CPU / memory / GPU telemetry for a running Slurm job — the facts, so you can judge.

Python 3.10+ MIT License

slurmwatch live TUI: per-process CPU / memory / GPU bars (each GPU's power shown against its cap), a JOB provenance card, and a wall-clock time-budget bar, with the MEM row lighting up amber then red as the working set climbs toward the OOM guard.

See whether your running job is actually using the CPU, memory, and GPUs you requested — live, in your terminal.

Install

pip install slurmwatch      # or: uv tool install slurmwatch / pipx install slurmwatch

Python 3.10+ on Linux (cgroup v1/v2). GPU monitoring auto-activates on NVIDIA nodes, and is skipped on CPU-only ones.

Usage

slurmwatch 12345    # watch a specific job
slurmwatch          # or auto-discover your running job
sw 12345            # "sw" is a short alias
slurmwatch --help   # everything else

Keys — c/m/g drill into CPU / memory / GPU · type a node number (or ←/→) to switch node · p expand a truncated path · q back/quit.

It counts only your PIDs (a neighbour on a shared node never inflates your numbers), tracks the real working set against your --mem, and flags an idle GPU that's still holding VRAM — the stuff nvidia-smi and htop won't tell you about your job.

Point it at a pending job and, instead of an error, you get why it's waiting, when it'll start and where you sit in line, and where it could run right now — with the exact scontrol update to requeue into a partition that fits.

Run it from a login node or on the node itself — it attaches either way, and falls back to an sstat summary when it can't.

Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Source Distribution

slurmwatch-1.1.0.tar.gz (4.2 MB view details)

Uploaded Jul 30, 2026 Source

Built Distribution

If you're not sure about the file name format, learn more about wheel file names.

The dropdown lists show the available interpreters, ABIs, and platforms. Enable javascript to be able to filter the list of wheel files.

slurmwatch-1.1.0-py3-none-any.whl (181.6 kB view details)

Uploaded Jul 30, 2026 Python 3

File details

Details for the file slurmwatch-1.1.0.tar.gz.

File metadata

Download URL: slurmwatch-1.1.0.tar.gz
Upload date: Jul 30, 2026
Size: 4.2 MB
Tags: Source
Uploaded using Trusted Publishing? Yes
Uploaded via: twine/7.0.0 CPython/3.13.14

File hashes

Hashes for slurmwatch-1.1.0.tar.gz
Algorithm	Hash digest
SHA256	`fb3d1a74e1e747ce74a3c207d43f91c0ac8cea60a3800a6b7d4ae7c9f4526c2f`
MD5	`aea39f032258ec3768dd3740bfd0ed6d`
BLAKE2b-256	`f65ac58e67eb69c60f2452db21a9c9c680615e6e3143917f0a56c2664d75283d`

See more details on using hashes here.

Provenance

The following attestation bundles were made for slurmwatch-1.1.0.tar.gz:

Publisher: release.yml on PursuitOfDataScience/slurmwatch

Attestations: Values shown here reflect the state when the release was signed and may no longer be current.

Statement:
- Statement type: https://in-toto.io/Statement/v1
- Predicate type: https://docs.pypi.org/attestations/publish/v1
- Subject name: slurmwatch-1.1.0.tar.gz
- Subject digest: fb3d1a74e1e747ce74a3c207d43f91c0ac8cea60a3800a6b7d4ae7c9f4526c2f
- Sigstore transparency entry: 2286358958
- Sigstore integration time: Jul 30, 2026
Source repository:
- Permalink: PursuitOfDataScience/slurmwatch@6fd011226d6587101cb44fda0979cebbb849711a
- Branch / Tag: refs/tags/v1.1.0
- Owner: https://github.com/PursuitOfDataScience
- Access: public
Publication detail:
- Token Issuer: https://token.actions.githubusercontent.com
- Runner Environment: github-hosted
- Publication workflow: release.yml@6fd011226d6587101cb44fda0979cebbb849711a
- Trigger Event: push