slurm-workload-manager

分类: 文档与知识管理 | 上传者: Lyken17 | 下载: 0 | 版本: v1.0（最新）

Comprehensive guide and documentation for Slurm Workload Manager. Use when submitting jobs, managing queues, configuring Slurm, or troubleshooting Slurm-managed clusters.

更新日志: Source: GitHub https://github.com/Lyken17/slurm-doc-markdown

目录结构

当前层级: slurm-workload-manager/

📁 references/
- 📄 accounting.md 42.5 KB
- 📄 acct_gather.conf.md 14.5 KB
- 📄 add.md 2.8 KB
- 📄 api.md 1.1 KB
- 📄 authentication.md 8.1 KB
- 📄 big_sys.md 9.3 KB
- 📄 burst_buffer.conf.md 7.0 KB
- 📄 burst_buffer.md 32.0 KB
- 📄 certmgr.md 8.1 KB
- 📄 cgroup.conf.md 13.5 KB
- 📄 cgroup_v2.md 33.3 KB
- 📄 cgroups.md 15.4 KB
- 📄 classic_fair_share.md 13.3 KB
- 📄 cli_filter_plugins.md 10.8 KB
- 📄 configless_slurm.md 6.3 KB
- 📄 configurator.easy.md 6.6 KB
- 📄 configurator.md 7.0 KB
- 📄 cons_tres.md 13.9 KB
- 📄 cons_tres_share.md 11.4 KB
- 📄 containers.md 33.8 KB
- 📄 contributor.md 3.9 KB
- 📄 core_spec.md 6.5 KB
- 📄 cpu_management.md 44.3 KB
- 📄 dist_plane.md 4.1 KB
- 📄 documentation.md 5.2 KB
- 📄 dynamic_nodes.md 3.5 KB
- 📄 elastic_computing.md 27.6 KB
- 📄 elasticsearch.md 5.8 KB
- 📄 extra_constraints.md 7.2 KB
- 📄 fair_tree.md 6.7 KB
- 📄 faq.md 101.4 KB
- 📄 federation.md 15.8 KB
- 📄 gang_scheduling.md 20.1 KB
- 📄 gres.conf.md 19.0 KB
- 📄 gres.md 33.0 KB
- 📄 gres_design.md 6.6 KB
- 📄 hdf5_profile_user_guide.md 11.9 KB
- 📄 helpers.conf.md 7.3 KB
- 📄 heterogeneous_jobs.md 29.9 KB
- 📄 high_throughput.md 16.2 KB
- 📄 hres.md 8.2 KB
- 📄 job_array.md 14.5 KB
- 📄 job_container.conf.md 8.6 KB
- 📄 job_exit_code.md 5.2 KB
- 📄 job_launch.md 9.2 KB
- 📄 job_reason_codes.md 27.0 KB
- 📄 job_state_codes.md 4.9 KB
- 📄 job_submit_plugins.md 13.5 KB
- 📄 jobcomp_kafka.md 7.1 KB
- 📄 jwt.md 14.2 KB
- 📄 kubernetes.md 1.8 KB
- 📄 licenses.md 14.2 KB
- 📄 mail.md 1.5 KB
- 📄 man_index.md 4.4 KB
- 📄 mc_support.md 26.6 KB
- 📄 mcs.md 9.0 KB
- 📄 metrics.md 10.3 KB
- 📄 mpi.conf.md 5.1 KB
- 📄 mpi_guide.md 25.4 KB
- 📄 multi_cluster.md 3.1 KB
- 📄 namespace.md 4.0 KB
- 📄 namespace.yaml.md 9.3 KB
- 📄 network.md 9.9 KB
- 📄 nss_slurm.md 6.9 KB
- 📄 oci.conf.md 7.6 KB
- 📄 openapi_release_notes.md 1.2 MB
- 📄 overview.md 9.2 KB
- 📄 pam_slurm_adopt.md 18.5 KB
- 📄 platforms.md 3.1 KB
- 📄 plugins.md 10.1 KB
- 📄 power_save.md 27.6 KB
- 📄 preempt.md 24.7 KB
- 📄 prep_plugins.md 5.2 KB
- 📄 priority_multifactor.md 20.3 KB
- 📄 priority_multifactor3.md 4.6 KB
- 📄 programmer_guide.md 13.2 KB
- 📄 prolog_epilog.md 18.6 KB
- 📄 qos.md 14.2 KB
- 📄 quickstart.md 15.5 KB
- 📄 quickstart_admin.md 33.5 KB
- 📄 related_software.md 9.6 KB
- 📄 release_notes.md 5.8 KB
- 📄 reservations.md 22.5 KB
- 📄 resource_binding.md 3.6 KB
- 📄 resource_limits.md 22.0 KB
- 📄 resources.yaml.md 6.0 KB
- 📄 rest.md 15.3 KB
- 📄 rest_api.md 498.5 KB
- 📄 rest_clients.md 16.6 KB
- 📄 rest_quickstart.md 12.2 KB
- 📄 rosetta.md 1.3 KB
- 📄 sacct.md 44.3 KB
- 📄 sacctmgr.md 97.3 KB
- 📄 sackd.md 5.5 KB
- 📄 salloc.md 100.7 KB
- 📄 sattach.md 5.7 KB
- 📄 sbatch.md 115.5 KB
- 📄 sbcast.md 7.7 KB
- 📄 scancel.md 12.8 KB
- 📄 sched_config.md 9.6 KB
- 📄 scontrol.md 109.3 KB
- 📄 scrontab.md 7.9 KB
- 📄 scrun.md 31.9 KB
- 📄 sdiag.md 13.7 KB
- 📄 select_design.md 4.4 KB
- 📄 selinux.md 5.3 KB
- 📄 sh5util.md 5.6 KB
- 📄 sinfo.md 29.8 KB
- 📄 site_factor.md 4.2 KB
- 📄 slinky.md 3.3 KB
- 📄 slurm.conf.md 284.4 KB
- 📄 slurm.md 6.2 KB
- 📄 slurmctld.md 8.0 KB
- 📄 slurmd.md 10.8 KB
- 📄 slurmdbd.conf.md 32.2 KB
- 📄 slurmdbd.md 4.8 KB
- 📄 slurmrestd.md 22.4 KB
- 📄 slurmstepd.md 2.9 KB
- 📄 spank.md 23.1 KB
- 📄 sprio.md 9.0 KB
- 📄 squeue.md 46.8 KB
- 📄 sreport.md 25.4 KB
- 📄 srun.md 152.3 KB
- 📄 sshare.md 8.7 KB
- 📄 sstat.md 9.7 KB
- 📄 strigger.md 14.2 KB
- 📄 sview.md 4.5 KB
- 📄 tls.md 5.4 KB
- 📄 topology.conf.md 6.2 KB
- 📄 topology.md 14.3 KB
- 📄 topology.yaml.md 4.9 KB
- 📄 tres.md 7.3 KB
- 📄 troubleshoot.md 9.1 KB
- 📄 upgrades.md 23.2 KB
- 📄 user_permissions.md 2.1 KB
- 📄 wckey.md 3.0 KB
📄 SKILL.md 3.2 KB

SKILL.md

---
name: slurm-workload-manager
description: Comprehensive guide and documentation for Slurm Workload Manager. Use when submitting jobs, managing queues, configuring Slurm, or troubleshooting Slurm-managed clusters.
---

# Slurm Workload Manager

This skill provides a complete reference for interacting with and managing a Slurm-based HPC cluster. It includes detailed command guides, job submission templates, and extensive configuration references.

## Core Workflows

### 1. Job Submission and Management
Submit and monitor your computational tasks.
- **Submit a batch script**: Use `sbatch`. See [sbatch.md](references/sbatch.md) for full options.
- **Run interactive tasks**: Use `srun`. See [srun.md](references/srun.md).
- **Check job status**: Use `squeue`. See [squeue.md](references/squeue.md).
- **Cancel jobs**: Use `scancel`. See [scancel.md](references/scancel.md).

### 2. Monitoring Cluster State
View nodes, partitions, and resource availability.
- **Cluster overview**: Use `sinfo`. See [sinfo.md](references/sinfo.md).
- **Node/Job details**: Use `scontrol show [node|job] <id>`. See [scontrol.md](references/scontrol.md).
- **Accounting & history**: Use `sacct`. See [sacct.md](references/sacct.md).

### 3. Cluster Administration
Manage Slurm configuration, users, and accounts.
- **Manage accounts/users**: Use `sacctmgr`. See [sacctmgr.md](references/sacctmgr.md).
- **Configuration files**: The primary configuration is [slurm.conf.md](references/slurm.conf.md).
- **Generic Resources (GRES)**: Configure GPUs and other resources in [gres.conf.md](references/gres.conf.md).

---

## Command Reference Library

Detailed documentation for all Slurm utilities:

| Command | Description | Reference |
| :--- | :--- | :--- |
| `sbatch` | Submit a batch script | [sbatch.md](references/sbatch.md) |
| `squeue` | View job queue | [squeue.md](references/squeue.md) |
| `srun` | Submit/run job steps | [srun.md](references/srun.md) |
| `scancel` | Cancel jobs | [scancel.md](references/scancel.md) |
| `sinfo` | View cluster state | [sinfo.md](references/sinfo.md) |
| `scontrol` | View/modify state | [scontrol.md](references/scontrol.md) |
| `sacct` | View job accounting | [sacct.md](references/sacct.md) |
| `sacctmgr` | Manage accounting | [sacctmgr.md](references/sacctmgr.md) |
| `salloc` | Allocate resources | [salloc.md](references/salloc.md) |
| `sreport` | Generate reports | [sreport.md](references/sreport.md) |

## Configuration Reference

Documentation for Slurm configuration files:

- **Main Config**: [slurm.conf.md](references/slurm.conf.md)
- **Database Config**: [slurmdbd.conf.md](references/slurmdbd.conf.md)
- **GRES (GPU) Config**: [gres.conf.md](references/gres.conf.md)
- **Cgroup Config**: [cgroup.conf.md](references/cgroup.conf.md)
- **Topology Config**: [topology.conf.md](references/topology.conf.md)

## Specialized Guides

- **Job Arrays**: [job_array.md](references/job_array.md)
- **QoS (Quality of Service)**: [qos.md](references/qos.md)
- **Reservations**: [reservations.md](references/reservations.md)
- **REST API**: [rest_api.md](references/rest_api.md)
- **Troubleshooting**: [troubleshoot.md](references/troubleshoot.md)
- **FAQs**: [faq.md](references/faq.md)

For any topic not listed above, use `grep` to search the `references/` directory for specific keywords.

登录后下载/点赞/收藏 ❤ 4 | ★ 0

slurm-workload-manager

目录结构

SKILL.md

举报内容

提示