# SchedMD/slurm

Slurm: A Highly Scalable Workload Manager

Repository: https://github.com/SchedMD/slurm
Canonical: https://ross.abutalabs.com/products/slurm
Homepage: https://slurm.schedmd.com/
Language: C
License: NOASSERTION
License Family: other
Topics: slurm, slurm-job-scheduler, slurm-workload-manager
Last push: 2026-08-26T21:20:03+00:00

## Health v2 (maintenance only)
Score: 98/100 (v2, computed 2026-09-03T02:20:16.233290+00:00)
- activity 99, release rhythm 97, longevity 100
- inputs: {"age_days": 5553, "days_push": 7, "days_rel": 20, "gap_med": 0.0, "n_releases_24m": 21}
- flags: no_license
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 4282, forks 881 (observed 2026-08-28T04:08:41.180669+00:00)

## What it is
Slurm is an open-source, highly scalable cluster resource management and job scheduling system for Linux clusters. It allocates compute nodes to users, launches and monitors parallel jobs, and arbitrates resource contention through a queue of pending work.

## Use cases
- schedule batch and parallel jobs on an HPC cluster
- manage a queue of pending jobs across shared compute nodes
- allocate exclusive or shared access to cluster resources for a time window
- run MPI workloads across many nodes
- implement fairshare and priority-based job scheduling
- expose a REST API for submitting and monitoring cluster jobs
- track cluster usage with accounting and reporting

## When to choose
- you operate a Linux cluster (HPC, research, GPU farm) needing centralized job scheduling
- you need scalable, fault-tolerant resource arbitration across hundreds or thousands of nodes
- you want fairshare, backfill, job arrays, and heterogeneous job support out of the box
- you need accounting, prioritization, and multi-cluster federation

## When to avoid
- you only need to orchestrate containers on a small set of machines - Kubernetes may fit better
- you need a lightweight single-machine task runner like cron or GNU parallel
- your infrastructure is not Linux - Slurm is only tested on Linux
- you want a managed cloud-native scheduler without cluster administration overhead

## Facets
- artifact type: application
- maturity: stable
- function: scheduling, workflow-automation, monitoring, cli, api-framework, auth
- domain: microservices, infrastructure-as-code, self-hosted
- platform: cpp, cli, self-hosted
- tags: job-scheduler, workload-manager, hpc, cluster-computing, batch-scheduling, resource-management, mpi, rest-api, fairshare, devops, high-performance-computing, linux

## Member repositories
- SchedMD/slurm (main) score 98

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:08:41.180669+00:00.
- Health v2: computed from the inputs above; adoption is never an input.
- Inferred fields (summary, facets, guidance): AI-extracted, prompt v1, taxonomy v1, on 2026-08-29T18:21:57.877902+00:00, confidence not recorded.
  - readme: https://github.com/SchedMD/slurm (fetched 2026-08-28T04:08:41.180669+00:00, sha 3f804dd01c88)
  - homepage: https://slurm.schedmd.com/ (fetched 2026-08-29T09:11:24.662588+00:00, sha 4256f9f38296)
  - site_page: https://slurm.schedmd.com/documentation.html (fetched 2026-08-29T09:11:24.666315+00:00, sha 4256f9f38296)
  - site_page: https://slurm.schedmd.com/quickstart_admin.html (fetched 2026-08-29T09:11:24.673153+00:00, sha b74710b4ed8f)
  - site_page: https://slurm.schedmd.com/quickstart.html (fetched 2026-08-29T09:11:24.675780+00:00, sha d827b3e07682)
  - site_page: https://slurm.schedmd.com/rest_quickstart.html (fetched 2026-08-29T09:11:24.677918+00:00, sha 5796e1c4669a)
  - site_page: https://slurm.schedmd.com/faq.html (fetched 2026-08-29T09:11:24.668349+00:00, sha 566435996f18)
- Data as of 2026-08-30T08:39:29.467469+00:00.