Prometheus MCP Server Extended
Enables querying real-time and historical server performance metrics (CPU, memory, disk IO, load, TCP) from Prometheus via Node Exporter.
README
Prometheus MCP Server Extended
基于 Prometheus 的节点性能指标查询服务,支持实时指标采集和历史趋势查询。
功能特性
- 实时指标查询: 传入节点名/IP,返回当前 CPU、内存、IO、负载、TCP 等实时指标
- 趋势数据查询: 传入节点名/IP + 时间范围,返回历史趋势数据
- 多指标支持: CPU、内存、磁盘 IO、系统负载、TCP 连接等
- RESTful API: 标准 REST 接口,支持 JSON 格式响应
- Swagger 文档: 自动生成 API 文档
快速开始
1. 安装依赖
cd prometheus-mcp-server
pip install -r requirements.txt
2. 配置环境变量
# 创建 .env 文件
cat > .env << EOF
PROMETHEUS_URL=http://localhost:9090
NODE_EXPORTER_PORT=9100
API_PORT=8000
EOF
3. 启动服务
# 开发模式
python -m app.main
# 或使用 uvicorn
uvicorn app.main:app --host 0.0.0.0 --port 8000 --reload
4. 访问 API 文档
- Swagger UI: http://localhost:8000/docs
- ReDoc: http://localhost:8000/redoc
API 接口
实时指标查询
GET /api/v1/metrics/realtime?node=node-01&ip=192.168.1.10
响应示例:
{
"node": "node-01",
"ip": "192.168.1.10",
"timestamp": "2026-06-25T12:00:00",
"cpu": {
"usage_percent": 25.5,
"cores": 4,
"user_percent": 10.0,
"system_percent": 5.0,
"iowait_percent": 2.0
},
"memory": {
"total_bytes": 8589934592,
"used_bytes": 4294967296,
"available_bytes": 4294967296,
"usage_percent": 50.0
},
"disk_io": {
"read_bytes_per_sec": 1024.0,
"write_bytes_per_sec": 512.0,
"read_iops": 10.0,
"write_iops": 5.0
},
"load": {
"load_1m": 2.5,
"load_5m": 3.0,
"load_15m": 3.5
},
"tcp": {
"connections_established": 100,
"listen_overflows": 0
}
}
趋势数据查询
GET /api/v1/metrics/trend?node=node-01&ip=192.168.1.10&start_time=2026-06-25T00:00:00Z&end_time=2026-06-25T23:59:59Z&step=5m
响应示例:
{
"node": "node-01",
"ip": "192.168.1.10",
"time_range": {
"start": "2026-06-25T00:00:00Z",
"end": "2026-06-25T23:59:59Z",
"step": "5m"
},
"cpu_trend": [
{"timestamp": "2026-06-25T00:00:00", "value": 25.5},
{"timestamp": "2026-06-25T00:05:00", "value": 26.0}
],
"memory_trend": [...],
"disk_io_trend": [...],
"load_trend": [...],
"tcp_trend": [...]
}
Docker 部署
# 构建镜像
docker build -t prometheus-mcp-server:latest .
# 运行容器
docker run -d \
--name prometheus-mcp-server \
-p 8000:8000 \
-e PROMETHEUS_URL=http://prometheus:9090 \
prometheus-mcp-server:latest
# 或使用 docker-compose
docker-compose up -d
Kubernetes 部署
# 应用配置
kubectl apply -f kubernetes/deployment.yaml
# 检查状态
kubectl get pods -l app=prometheus-mcp-server
kubectl get services
前置条件
- Prometheus: 需要运行 Prometheus 服务(默认端口 9090)
- Node Exporter: 目标节点需要部署 Node Exporter(默认端口 9100)
Node Exporter 安装
# Docker 方式
docker run -d \
--name node-exporter \
-p 9100:9100 \
-v /proc:/host/proc:ro \
-v /sys:/host/sys:ro \
prom/node-exporter \
--path.procfs=/host/proc \
--path.sysfs=/host/sys
# 或直接安装
wget https://github.com/prometheus/node_exporter/releases/download/v1.6.0/node_exporter-1.6.0.linux-amd64.tar.gz
tar xzf node_exporter-*.tar.gz
./node_exporter
配置说明
| 参数 | 默认值 | 说明 |
|---|---|---|
PROMETHEUS_URL |
http://localhost:9090 |
Prometheus 服务地址 |
NODE_EXPORTER_PORT |
9100 |
Node Exporter 端口 |
API_PORT |
8000 |
API 服务端口 |
DEFAULT_STEP |
1m |
默认采样间隔 |
REALTIME_CACHE_TTL |
30 |
实时数据缓存时间(秒) |
测试
# 运行测试
pytest tests/ -v
# 或使用 unittest
python -m pytest tests/test_services.py -v
项目结构
prometheus-mcp-server/
├── app/
│ ├── api/
│ │ └── routes.py # API 路由
│ ├── services/
│ │ ├── realtime_collector.py # 实时采集服务
│ │ └── trend_query_service.py # 趋势查询服务
│ ├── models/
│ │ └── schemas.py # 数据模型
│ ├── config/
│ │ └── settings.py # 配置管理
│ └── main.py # FastAPI 入口
├── tests/
│ └── test_services.py # 测试文件
├── kubernetes/
│ └── deployment.yaml # Kubernetes 配置
├── Dockerfile # Docker 镜像配置
├── docker-compose.yml # Docker Compose 配置
├── prometheus.yml # Prometheus 配置
├── requirements.txt # Python 依赖
└── README.md # 项目说明
许可证
MIT License
Recommended Servers
playwright-mcp
A Model Context Protocol server that enables LLMs to interact with web pages through structured accessibility snapshots without requiring vision models or screenshots.
Magic Component Platform (MCP)
An AI-powered tool that generates modern UI components from natural language descriptions, integrating with popular IDEs to streamline UI development workflow.
Audiense Insights MCP Server
Enables interaction with Audiense Insights accounts via the Model Context Protocol, facilitating the extraction and analysis of marketing insights and audience data including demographics, behavior, and influencer engagement.
VeyraX MCP
Single MCP tool to connect all your favorite tools: Gmail, Calendar and 40 more.
graphlit-mcp-server
The Model Context Protocol (MCP) Server enables integration between MCP clients and the Graphlit service. Ingest anything from Slack to Gmail to podcast feeds, in addition to web crawling, into a Graphlit project - and then retrieve relevant contents from the MCP client.
Kagi MCP Server
An MCP server that integrates Kagi search capabilities with Claude AI, enabling Claude to perform real-time web searches when answering questions that require up-to-date information.
E2B
Using MCP to run code via e2b.
Neon Database
MCP server for interacting with Neon Management API and databases
Exa Search
A Model Context Protocol (MCP) server lets AI assistants like Claude use the Exa AI Search API for web searches. This setup allows AI models to get real-time web information in a safe and controlled way.
Qdrant Server
This repository is an example of how to create a MCP server for Qdrant, a vector search engine.