mirror of https://github.com/nicolargo/glances.git
Add information about local tesbed with Ollama + WebUI
This commit is contained in:
nicolargo
parent
22e90ee9dd
commit
c1bfe0b063
|
|
@ -0,0 +1,202 @@
|
|||
# Glances MCP Server Testing Setup
|
||||
|
||||
## Prerequisites
|
||||
|
||||
```bash
|
||||
# Install Docker and Docker Compose
|
||||
sudo apt update
|
||||
sudo apt install docker.io docker-compose
|
||||
sudo usermod -aG docker $USER
|
||||
# Log out and back in for group changes to take effect
|
||||
```
|
||||
|
||||
## Setup Steps
|
||||
|
||||
### 1. Start the Services
|
||||
|
||||
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:
|
||||
docker compose up -d
|
||||
|
||||
# Check if services are running:
|
||||
docker compose ps
|
||||
```
|
||||
|
||||
### 2. Pull an LLM Model
|
||||
|
||||
```bash
|
||||
# Pull a recommended model (choose based on your RAM):
|
||||
|
||||
# For 8GB RAM - Llama 3.2 (3B parameters)
|
||||
docker exec -it ollama ollama pull llama3.2
|
||||
|
||||
# For 16GB RAM - Llama 3.1 (8B parameters)
|
||||
docker exec -it ollama ollama pull llama3.1
|
||||
|
||||
# For 32GB+ RAM - Llama 3.1 (70B parameters)
|
||||
docker exec -it ollama ollama pull llama3.1:70b
|
||||
|
||||
# List available models:
|
||||
docker exec -it ollama ollama list
|
||||
```
|
||||
|
||||
### 3. Access the Services
|
||||
|
||||
- **Open WebUI**: http://localhost:3000
|
||||
- Create an account on first visit
|
||||
- The LLM will be automatically detected
|
||||
|
||||
- **Glances API**: http://localhost:61208
|
||||
- API endpoint: http://localhost:61208/api/4/
|
||||
- Web interface: http://localhost:61208
|
||||
|
||||
### 4. Test the Glances API
|
||||
|
||||
```bash
|
||||
# Test basic endpoints:
|
||||
curl http://localhost:61208/api/4/status
|
||||
curl http://localhost:61208/api/4/pluginslist
|
||||
curl http://localhost:61208/api/4/cpu
|
||||
curl http://localhost:61208/api/4/mem
|
||||
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:
|
||||
|
||||
**Example MCP Configuration** (adjust based on your MCP implementation):
|
||||
|
||||
```json
|
||||
{
|
||||
"mcpServers": {
|
||||
"glances": {
|
||||
"command": "node",
|
||||
"args": ["/path/to/your/glances-mcp-server/index.js"],
|
||||
"env": {
|
||||
"GLANCES_API_URL": "http://host.docker.internal:61208/api/4"
|
||||
}
|
||||
}
|
||||
}
|
||||
}
|
||||
```
|
||||
|
||||
**Note**: Use `host.docker.internal` to access services on your host machine from within Docker containers.
|
||||
|
||||
## Intel GPU Acceleration (Optional)
|
||||
|
||||
If you want to enable Intel GPU acceleration:
|
||||
|
||||
### 1. Install Intel GPU Drivers
|
||||
|
||||
```bash
|
||||
# Install Intel compute runtime
|
||||
sudo apt update
|
||||
sudo apt install -y intel-opencl-icd intel-media-va-driver-non-free
|
||||
|
||||
# Verify GPU is accessible
|
||||
ls -la /dev/dri
|
||||
```
|
||||
|
||||
### 2. Enable GPU in Docker Compose
|
||||
|
||||
Uncomment these lines in `docker-compose.yml`:
|
||||
|
||||
```yaml
|
||||
devices:
|
||||
- /dev/dri:/dev/dri
|
||||
environment:
|
||||
- OLLAMA_GPU_DRIVER=intel
|
||||
```
|
||||
|
||||
Then restart:
|
||||
|
||||
```bash
|
||||
docker-compose down
|
||||
docker-compose up -d
|
||||
```
|
||||
|
||||
## Testing Your MCP Server
|
||||
|
||||
### Manual Test with cURL
|
||||
|
||||
```bash
|
||||
# Test if your MCP server can call Glances endpoints
|
||||
# Example: Get CPU stats
|
||||
curl http://localhost:61208/api/4/cpu
|
||||
|
||||
# Expected response: JSON with CPU stats
|
||||
```
|
||||
|
||||
### Test in Open WebUI
|
||||
|
||||
1. Open http://localhost:3000
|
||||
2. Start a new chat
|
||||
3. Try prompts like:
|
||||
- "What is the current CPU usage?"
|
||||
- "Show me memory statistics"
|
||||
- "List all available system plugins"
|
||||
|
||||
## Useful Commands
|
||||
|
||||
```bash
|
||||
# View logs
|
||||
docker-compose logs -f ollama
|
||||
docker-compose logs -f open-webui
|
||||
docker-compose logs -f glances
|
||||
|
||||
# Restart services
|
||||
docker-compose restart
|
||||
|
||||
# Stop all services
|
||||
docker-compose down
|
||||
|
||||
# Stop and remove volumes (clean start)
|
||||
docker-compose down -v
|
||||
|
||||
# Check Ollama models
|
||||
docker exec -it ollama ollama list
|
||||
|
||||
# Monitor resource usage
|
||||
docker stats
|
||||
```
|
||||
|
||||
## Troubleshooting
|
||||
|
||||
### Ollama not responding
|
||||
```bash
|
||||
docker-compose restart ollama
|
||||
docker exec -it ollama ollama list
|
||||
```
|
||||
|
||||
### Glances API not accessible
|
||||
```bash
|
||||
# Check if Glances is running
|
||||
curl http://localhost:61208/api/4/status
|
||||
|
||||
# View Glances logs
|
||||
docker-compose logs glances
|
||||
```
|
||||
|
||||
### Low memory issues
|
||||
```bash
|
||||
# Use a smaller model
|
||||
docker exec -it ollama ollama pull llama3.2:1b
|
||||
|
||||
# Or configure Ollama to use less memory
|
||||
docker-compose down
|
||||
# Add to ollama service environment:
|
||||
# - OLLAMA_MAX_LOADED_MODELS=1
|
||||
# - OLLAMA_NUM_PARALLEL=1
|
||||
docker-compose up -d
|
||||
```
|
||||
|
||||
## Recommended Models for Your Hardware
|
||||
|
||||
- **4-8GB RAM**: `llama3.2:1b` or `phi3:mini`
|
||||
- **8-16GB RAM**: `llama3.2` or `mistral`
|
||||
- **16GB+ RAM**: `llama3.1` or `mixtral`
|
||||
|
||||
Choose smaller models for better responsiveness on limited hardware.
|
||||
|
|
@ -0,0 +1,52 @@
|
|||
version: '3.8'
|
||||
|
||||
services:
|
||||
# Ollama - Local LLM Runtime
|
||||
ollama:
|
||||
image: ollama/ollama:latest
|
||||
container_name: ollama
|
||||
ports:
|
||||
- "11434:11434"
|
||||
volumes:
|
||||
- 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
|
||||
|
||||
# Open WebUI - Web Interface for Ollama
|
||||
open-webui:
|
||||
image: ghcr.io/open-webui/open-webui:main
|
||||
container_name: open-webui
|
||||
ports:
|
||||
- "3000:8080"
|
||||
environment:
|
||||
- OLLAMA_BASE_URL=http://ollama:11434
|
||||
- WEBUI_SECRET_KEY=your-secret-key-change-this
|
||||
volumes:
|
||||
- open_webui_data:/app/backend/data
|
||||
depends_on:
|
||||
- ollama
|
||||
restart: unless-stopped
|
||||
extra_hosts:
|
||||
- "host.docker.internal:host-gateway"
|
||||
|
||||
# Glances - System Monitoring (with API enabled)
|
||||
glances:
|
||||
image: nicolargo/glances:latest
|
||||
container_name: glances
|
||||
ports:
|
||||
- "61208:61208"
|
||||
environment:
|
||||
- GLANCES_OPT=-w
|
||||
volumes:
|
||||
- /var/run/docker.sock:/var/run/docker.sock:ro
|
||||
- /etc/os-release:/etc/os-release:ro
|
||||
pid: host
|
||||
restart: unless-stopped
|
||||
|
||||
volumes:
|
||||
ollama_data:
|
||||
open_webui_data:
|
||||
|
|
@ -0,0 +1,154 @@
|
|||
#!/bin/bash
|
||||
|
||||
# Colors for output
|
||||
GREEN='\033[0;32m'
|
||||
RED='\033[0;31m'
|
||||
YELLOW='\033[1;33m'
|
||||
NC='\033[0m' # No Color
|
||||
|
||||
echo "=========================================="
|
||||
echo " Glances MCP Server Setup Verification"
|
||||
echo "=========================================="
|
||||
echo ""
|
||||
|
||||
# Check if Docker is running
|
||||
echo -n "Checking Docker... "
|
||||
if docker info > /dev/null 2>&1; then
|
||||
echo -e "${GREEN}✓${NC}"
|
||||
else
|
||||
echo -e "${RED}✗${NC}"
|
||||
echo "Error: Docker is not running. Please start Docker and try again."
|
||||
exit 1
|
||||
fi
|
||||
|
||||
# Check if services are running
|
||||
echo -n "Checking Ollama service... "
|
||||
if docker ps | grep -q ollama; then
|
||||
echo -e "${GREEN}✓${NC}"
|
||||
else
|
||||
echo -e "${RED}✗${NC}"
|
||||
echo "Error: Ollama container is not running. Run 'docker-compose up -d'"
|
||||
exit 1
|
||||
fi
|
||||
|
||||
echo -n "Checking Open WebUI service... "
|
||||
if docker ps | grep -q open-webui; then
|
||||
echo -e "${GREEN}✓${NC}"
|
||||
else
|
||||
echo -e "${RED}✗${NC}"
|
||||
echo "Warning: Open WebUI container is not running."
|
||||
fi
|
||||
|
||||
echo -n "Checking Glances service... "
|
||||
if docker ps | grep -q glances; then
|
||||
echo -e "${GREEN}✓${NC}"
|
||||
else
|
||||
echo -e "${RED}✗${NC}"
|
||||
echo "Warning: Glances container is not running."
|
||||
fi
|
||||
|
||||
echo ""
|
||||
echo "Testing API endpoints..."
|
||||
echo ""
|
||||
|
||||
# Test Ollama
|
||||
echo -n "Testing Ollama API... "
|
||||
if curl -s http://localhost:11434/api/tags > /dev/null 2>&1; then
|
||||
echo -e "${GREEN}✓${NC}"
|
||||
echo " Available models:"
|
||||
curl -s http://localhost:11434/api/tags | grep -o '"name":"[^"]*"' | cut -d'"' -f4 | sed 's/^/ - /'
|
||||
else
|
||||
echo -e "${RED}✗${NC}"
|
||||
echo " Error: Cannot connect to Ollama API"
|
||||
fi
|
||||
|
||||
echo ""
|
||||
|
||||
# Test Glances
|
||||
echo -n "Testing Glances API... "
|
||||
if curl -s http://localhost:61208/api/4/status > /dev/null 2>&1; then
|
||||
echo -e "${GREEN}✓${NC}"
|
||||
# Test specific endpoints
|
||||
echo " Testing endpoints:"
|
||||
echo -n " - /api/4/pluginslist... "
|
||||
if curl -s http://localhost:61208/api/4/pluginslist | grep -q '\['; then
|
||||
echo -e "${GREEN}✓${NC}"
|
||||
else
|
||||
echo -e "${RED}✗${NC}"
|
||||
fi
|
||||
echo -n " - /api/4/cpu... "
|
||||
if curl -s http://localhost:61208/api/4/cpu | grep -q 'total'; then
|
||||
echo -e "${GREEN}✓${NC}"
|
||||
else
|
||||
echo -e "${RED}✗${NC}"
|
||||
fi
|
||||
echo -n " - /api/4/mem... "
|
||||
if curl -s http://localhost:61208/api/4/mem | grep -q 'total'; then
|
||||
echo -e "${GREEN}✓${NC}"
|
||||
else
|
||||
echo -e "${RED}✗${NC}"
|
||||
fi
|
||||
else
|
||||
echo -e "${RED}✗${NC}"
|
||||
echo " Error: Cannot connect to Glances API"
|
||||
fi
|
||||
|
||||
echo ""
|
||||
|
||||
# Test Open WebUI
|
||||
echo -n "Testing Open WebUI... "
|
||||
if curl -s http://localhost:3000 > /dev/null 2>&1; then
|
||||
echo -e "${GREEN}✓${NC}"
|
||||
else
|
||||
echo -e "${RED}✗${NC}"
|
||||
echo " Error: Cannot connect to Open WebUI"
|
||||
fi
|
||||
|
||||
echo ""
|
||||
echo "=========================================="
|
||||
echo " Access URLs"
|
||||
echo "=========================================="
|
||||
echo "Open WebUI: http://localhost:3000"
|
||||
echo "Glances Web: http://localhost:61208"
|
||||
echo "Glances API: http://localhost:61208/api/4"
|
||||
echo "Ollama API: http://localhost:11434"
|
||||
echo ""
|
||||
|
||||
# Check for models
|
||||
echo "=========================================="
|
||||
echo " Ollama Models Status"
|
||||
echo "=========================================="
|
||||
if docker exec ollama ollama list > /dev/null 2>&1; then
|
||||
MODEL_COUNT=$(docker exec ollama ollama list | tail -n +2 | wc -l)
|
||||
if [ "$MODEL_COUNT" -eq 0 ]; then
|
||||
echo -e "${YELLOW}No models installed yet.${NC}"
|
||||
echo ""
|
||||
echo "To install a model, run:"
|
||||
echo " docker exec -it ollama ollama pull llama3.2"
|
||||
else
|
||||
echo -e "${GREEN}Installed models:${NC}"
|
||||
docker exec ollama ollama list
|
||||
fi
|
||||
else
|
||||
echo -e "${RED}Cannot check models${NC}"
|
||||
fi
|
||||
|
||||
echo ""
|
||||
echo "=========================================="
|
||||
echo " System Resources"
|
||||
echo "=========================================="
|
||||
echo "Container resource usage:"
|
||||
docker stats --no-stream --format "table {{.Container}}\t{{.CPUPerc}}\t{{.MemUsage}}"
|
||||
|
||||
echo ""
|
||||
echo "=========================================="
|
||||
echo " Next Steps"
|
||||
echo "=========================================="
|
||||
echo "1. Open WebUI at http://localhost:3000"
|
||||
echo "2. Create an account (stored locally)"
|
||||
echo "3. If no models, install one:"
|
||||
echo " docker exec -it ollama ollama pull llama3.2"
|
||||
echo "4. Test the Glances API in chat:"
|
||||
echo " - 'What is the current CPU usage?'"
|
||||
echo " - 'Show me memory statistics'"
|
||||
echo ""
|
||||
Loading…
Reference in New Issue