mirror of https://github.com/nicolargo/glances.git
Add MCPO local test
This commit is contained in:
nicolargo
parent
c1bfe0b063
commit
e9200db776
|
|
@ -14,7 +14,7 @@ sudo usermod -aG docker $USER
|
|||
|
||||
### 1. Start the Services
|
||||
|
||||
Warning: you should use docker-compose v2. If you have v1 installed, please uninstall it first.
|
||||
Warning: you should use docker compose v2. If you have v1 installed, please uninstall it first.
|
||||
|
||||
```bash
|
||||
# Save the docker-compose.yml file and run:
|
||||
|
|
@ -65,25 +65,92 @@ curl http://localhost:61208/api/4/all
|
|||
|
||||
### 5. Configure MCP Server Connection
|
||||
|
||||
In Open WebUI, you'll need to configure the MCP server. Here's how to test if your Glances MCP server can access the API:
|
||||
Install mcpo:
|
||||
|
||||
**Example MCP Configuration** (adjust based on your MCP implementation):
|
||||
```bash
|
||||
uv add mcpo
|
||||
```
|
||||
|
||||
```json
|
||||
Create MCPO configuration file:
|
||||
|
||||
```bash
|
||||
cd /home/nicolargo/dev/glances/mcp
|
||||
|
||||
# Create configuration
|
||||
cat > mcp_config.json << 'EOF'
|
||||
{
|
||||
"mcpServers": {
|
||||
"glances": {
|
||||
"command": "node",
|
||||
"args": ["/path/to/your/glances-mcp-server/index.js"],
|
||||
"command": "python3",
|
||||
"args": ["glances_mcp.py"],
|
||||
"env": {
|
||||
"GLANCES_API_URL": "http://host.docker.internal:61208/api/4"
|
||||
"GLANCES_API_URL": "http://localhost:61208/api/4"
|
||||
}
|
||||
}
|
||||
}
|
||||
}
|
||||
EOF
|
||||
```
|
||||
|
||||
**Note**: Use `host.docker.internal` to access services on your host machine from within Docker containers.
|
||||
Run MCPO proxy:
|
||||
|
||||
```bash
|
||||
../.venv-uv/bin/uv run mcpo --port 8000 --config mcp_config.json
|
||||
Starting MCP OpenAPI Proxy with config file: mcp_config.json
|
||||
2025-12-29 09:38:29,310 - INFO - Starting MCPO Server...
|
||||
2025-12-29 09:38:29,310 - INFO - Name: MCP OpenAPI Proxy
|
||||
2025-12-29 09:38:29,310 - INFO - Version: 1.0
|
||||
2025-12-29 09:38:29,310 - INFO - Description: Automatically generated API from MCP Tool Schemas
|
||||
2025-12-29 09:38:29,310 - INFO - Hostname: nicolargo-xps15
|
||||
2025-12-29 09:38:29,310 - INFO - Port: 8000
|
||||
2025-12-29 09:38:29,310 - INFO - API Key: Not Provided
|
||||
2025-12-29 09:38:29,310 - INFO - CORS Allowed Origins: ['*']
|
||||
2025-12-29 09:38:29,310 - INFO - Path Prefix: /
|
||||
2025-12-29 09:38:29,310 - INFO - Root Path:
|
||||
2025-12-29 09:38:29,311 - INFO - Loading MCP server configurations from: mcp_config.json
|
||||
2025-12-29 09:38:29,311 - INFO - Configuring MCP Servers:
|
||||
2025-12-29 09:38:29,311 - INFO - Uvicorn server starting...
|
||||
INFO: Started server process [167852]
|
||||
INFO: Waiting for application startup.
|
||||
2025-12-29 09:38:29,324 - INFO - Initiating connection for server: 'glances'...
|
||||
2025-12-29 09:38:29,800 - INFO - Successfully connected to 'glances'.
|
||||
2025-12-29 09:38:29,800 - INFO - --------------------------
|
||||
|
||||
INFO: Application startup complete.
|
||||
INFO: Uvicorn running on http://0.0.0.0:8000 (Press CTRL+C to quit)
|
||||
```
|
||||
|
||||
Check:
|
||||
|
||||
```bash
|
||||
curl http://localhost:8000/glances/openapi.json
|
||||
```
|
||||
|
||||
And configure Open WebUI to use MCP server at `http://host.docker.internal:8000/glances`.
|
||||
|
||||
Open Open WebUI: Go to `http://localhost:3000`
|
||||
Open Settings:
|
||||
|
||||
Click your profile icon (top right)
|
||||
Select "Settings"
|
||||
|
||||
Add Tool Server:
|
||||
|
||||
Navigate to "Tools" section in the left sidebar
|
||||
Click the "+" button to add a new tool server
|
||||
|
||||
Configure the Connection:
|
||||
|
||||
Server URL: `http://172.17.0.1:8000/glances`
|
||||
API Key: Leave blank (unless you configured one)
|
||||
|
||||
Enable the Tool:
|
||||
|
||||
The Glances tools should now appear in your tools list
|
||||
|
||||
In the conversation interface, click the "Tools" button (usually represented by a toolbox icon) and enable the Glances tool.
|
||||
|
||||
Test a prompt like: "What is the current CPU usage on my system ?"
|
||||
|
||||
## Intel GPU Acceleration (Optional)
|
||||
|
||||
|
|
@ -114,8 +181,8 @@ environment:
|
|||
Then restart:
|
||||
|
||||
```bash
|
||||
docker-compose down
|
||||
docker-compose up -d
|
||||
docker compose down
|
||||
docker compose up -d
|
||||
```
|
||||
|
||||
## Testing Your MCP Server
|
||||
|
|
@ -143,18 +210,18 @@ curl http://localhost:61208/api/4/cpu
|
|||
|
||||
```bash
|
||||
# View logs
|
||||
docker-compose logs -f ollama
|
||||
docker-compose logs -f open-webui
|
||||
docker-compose logs -f glances
|
||||
docker compose logs -f ollama
|
||||
docker compose logs -f open-webui
|
||||
docker compose logs -f glances
|
||||
|
||||
# Restart services
|
||||
docker-compose restart
|
||||
docker compose restart
|
||||
|
||||
# Stop all services
|
||||
docker-compose down
|
||||
docker compose down
|
||||
|
||||
# Stop and remove volumes (clean start)
|
||||
docker-compose down -v
|
||||
docker compose down -v
|
||||
|
||||
# Check Ollama models
|
||||
docker exec -it ollama ollama list
|
||||
|
|
@ -167,7 +234,7 @@ docker stats
|
|||
|
||||
### Ollama not responding
|
||||
```bash
|
||||
docker-compose restart ollama
|
||||
docker compose restart ollama
|
||||
docker exec -it ollama ollama list
|
||||
```
|
||||
|
||||
|
|
@ -177,7 +244,7 @@ docker exec -it ollama ollama list
|
|||
curl http://localhost:61208/api/4/status
|
||||
|
||||
# View Glances logs
|
||||
docker-compose logs glances
|
||||
docker compose logs glances
|
||||
```
|
||||
|
||||
### Low memory issues
|
||||
|
|
@ -186,11 +253,11 @@ docker-compose logs glances
|
|||
docker exec -it ollama ollama pull llama3.2:1b
|
||||
|
||||
# Or configure Ollama to use less memory
|
||||
docker-compose down
|
||||
docker compose down
|
||||
# Add to ollama service environment:
|
||||
# - OLLAMA_MAX_LOADED_MODELS=1
|
||||
# - OLLAMA_NUM_PARALLEL=1
|
||||
docker-compose up -d
|
||||
docker compose up -d
|
||||
```
|
||||
|
||||
## Recommended Models for Your Hardware
|
||||
|
|
|
|||
|
|
@ -11,10 +11,10 @@ services:
|
|||
- ollama_data:/root/.ollama
|
||||
restart: unless-stopped
|
||||
# Uncomment below for Intel GPU support (requires Intel GPU drivers on host)
|
||||
# devices:
|
||||
# - /dev/dri:/dev/dri
|
||||
# environment:
|
||||
# - OLLAMA_GPU_DRIVER=intel
|
||||
devices:
|
||||
- /dev/dri:/dev/dri
|
||||
environment:
|
||||
- OLLAMA_GPU_DRIVER=intel
|
||||
|
||||
# Open WebUI - Web Interface for Ollama
|
||||
open-webui:
|
||||
|
|
|
|||
|
|
@ -0,0 +1,11 @@
|
|||
{
|
||||
"mcpServers": {
|
||||
"glances": {
|
||||
"command": "python3",
|
||||
"args": ["glances_mcp.py"],
|
||||
"env": {
|
||||
"GLANCES_API_URL": "http://localhost:61208/api/4"
|
||||
}
|
||||
}
|
||||
}
|
||||
}
|
||||
|
|
@ -10,7 +10,6 @@ classifiers = [
|
|||
"Intended Audience :: System Administrators",
|
||||
"Operating System :: OS Independent",
|
||||
"Programming Language :: Python :: 3",
|
||||
"Programming Language :: Python :: 3.10",
|
||||
"Programming Language :: Python :: 3.11",
|
||||
"Programming Language :: Python :: 3.12",
|
||||
"Programming Language :: Python :: 3.13",
|
||||
|
|
@ -24,6 +23,9 @@ dependencies = [
|
|||
"windows-curses; platform_system == 'Windows'",
|
||||
"shtab; platform_system != 'Windows'",
|
||||
"jinja2",
|
||||
"mcpo>=0.0.19",
|
||||
"mcp>=1.25.0",
|
||||
"fastmcp>=2.14.1",
|
||||
]
|
||||
description = "A cross-platform curses-based monitoring tool"
|
||||
dynamic = ["version"]
|
||||
|
|
@ -31,7 +33,7 @@ keywords = ["cli", "curses", "monitoring", "system"]
|
|||
license = "LGPL-3.0-only"
|
||||
name = "Glances"
|
||||
readme = "README-pypi.rst"
|
||||
requires-python = ">=3.10"
|
||||
requires-python = ">=3.11"
|
||||
urls.Homepage = "https://github.com/nicolargo/glances"
|
||||
|
||||
[dependency-groups]
|
||||
|
|
|
|||
Loading…
Reference in New Issue