yidong72/slurm_mcp

Developer tools

Эта страница переведена автоматически. Технические детали смотрите в оригинале. Открыть оригинал

20 stars 3 forks Качество 63 Тренд 10

Repository Install

An MCP (Model Context Protocol) server that enables AI agents to interact with remote Slurm clusters via SSH.

Обзор

An MCP (Model Context Protocol) server that enables AI agents to interact with remote Slurm clusters via SSH. This server provides tools for job management, cluster monitoring, interactive sessions, and file operations. - : Submit, monitor, and cancel jobs via SSH - : Query GPU availability and allocate GPU resources - : Support for containerized workloads with .sqsh images - : Launch and manage persistent interactive sessions - : Save and reuse interactive session configurations - : Browse, read, and write files on the cluster - : Organized access to datasets, models, results, and logs The server is configured via environment variables with the SLURM_ prefix. The slurm-mcp-config tool automatically converts your .env file to mcp.json format: Options: - -o, --output FILE - Write to file instead of stdout - --merge FILE - Merge into an existing mcp.

README

Показан фрагмент оригинального README. Открыть оригинал.

Slurm MCP Server

Features

Remote Slurm Management: Submit, monitor, and cancel jobs via SSH
GPU Support: Query GPU availability and allocate GPU resources
Pyxis/Enroot Containers: Support for containerized workloads with .sqsh images
Interactive Sessions: Launch and manage persistent interactive sessions
Session Profiles: Save and reuse interactive session configurations
File Operations: Browse, read, and write files on the cluster
Directory Structure: Organized access to datasets, models, results, and logs

Installation

# Clone the repository
git clone https://github.com/yourusername/slurm-mcp.git
cd slurm-mcp

# Install with pip
pip install -e .

# Or install with development dependencies
pip install -e ".[dev]"

Quick Start

cd /path/to/slurm_mcp
pip install -e .

# Configure
cp .env.example .env
# Edit .env with your cluster details

# Generate mcp.json for Cursor/Claude Desktop
slurm-mcp-config .env -o mcp.json

# Or run directly
slurm-mcp

Configuration

The server is configured via environment variables with the SLURM_ prefix.

Required Settings

Variable	Description
`SLURM_SSH_HOST`	Remote Slurm login node hostname
`SLURM_SSH_USER`	SSH username
`SLURM_USER_ROOT`	User’s root directory on cluster

SSH Settings

Variable	Default	Description
`SLURM_SSH_PORT`	22	SSH port
`SLURM_SSH_KEY_PATH`	None	Path to SSH private key
`SLURM_SSH_PASSWORD`	None	SSH password or key passphrase
`SLURM_SSH_KNOWN_HOSTS`	None	Path to known_hosts file

Slurm Settings

Variable	Default	Description
`SLURM_DEFAULT_PARTITION`	None	Default partition for jobs
`SLURM_DEFAULT_ACCOUNT`	None	Default account/project
`SLURM_COMMAND_TIMEOUT`	60	Command timeout in seconds
`SLURM_GPU_PARTITIONS`	None	Comma-separated GPU partition names
`SLURM_CPU_PARTITIONS`	None	Comma-separated CPU partition names

Interactive Session Settings

Variable	Default	Description
`SLURM_INTERACTIVE_PARTITION`	interactive	Partition for interactive jobs
`SLURM_INTERACTIVE_ACCOUNT`	None	Account for interactive jobs
`SLURM_INTERACTIVE_DEFAULT_TIME`	4:00:00	Default time limit
`SLURM_INTERACTIVE_DEFAULT_GPUS`	8	Default GPUs per node
`SLURM_INTERACTIVE_SESSION_TIMEOUT`	3600	Idle timeout in seconds

Directory Structure

Variable	Default	Container Mount
`SLURM_USER_ROOT`	Required	-
`SLURM_DIR_DATASETS`	`$USER_ROOT/data`	`/datasets`

Открыть полный README на GitHub

Установка

This server does not publish a one-line install command.

Open the repository installation guide Новичок в MCP? Начните с руководства на mcpskill.org →

Конфигурация

{
  "mcpServers": {
    "slurm": {
      "command": "slurm-mcp",
      "env": {
        "SLURM_SSH_HOST": "login.cluster.example.com",
        "SLURM_SSH_USER": "username",
        "SLURM_SSH_KEY_PATH": "~/.ssh/id_rsa",
        "SLURM_USER_ROOT": "/lustre/users/username",
        "SLURM_GPFS_ROOT": "/lustre",
        "SLURM_DEFAULT_ACCOUNT": "my_project",
        "SLURM_DEFAULT_PARTITION": "batch",
        "SLURM_INTERACTIVE_PARTITION": "interactive"
      }
    }
  }
}