tony/hive
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity
Files
8619b75296b531d5de52705f0753e979a85b8f6a
hive/mcp-server/SERVICE_README.md
anthonyrawlins fc0eec91ef Complete Hive platform functionality and expand cluster to 7 agents 
Major Features Added:
- Fix Socket.IO connectivity by updating Dockerfile to use socket_app
- Resolve distributed workflows API to return arrays instead of errors
- Expand agent coverage from 3 to 7 agents (added OAK and ROSEWOOD)
- Create comprehensive systemd service for MCP server with auto-discovery
- Add daemon mode with periodic agent discovery every 5 minutes
- Implement comprehensive test suite with 100% pass rate

Infrastructure Improvements:
- Enhanced database connection handling with retry logic
- Improved agent registration with persistent storage
- Added proper error handling for distributed workflows endpoint
- Created management scripts for service lifecycle operations

Agent Cluster Expansion:
- ACACIA: deepseek-r1:7b (kernel_dev)
- WALNUT: starcoder2:15b (pytorch_dev)
- IRONWOOD: deepseek-coder-v2 (profiler)
- OAK: codellama:latest (docs_writer)
- OAK-TESTER: deepseek-r1:latest (tester)
- ROSEWOOD: deepseek-coder-v2:latest (kernel_dev)
- ROSEWOOD-VISION: llama3.2-vision:11b (tester)

System Status: All 7 agents healthy, Socket.IO operational, MCP server fully functional

🤖 Generated with [Claude Code](https://claude.ai/code)

Co-Authored-By: Claude <noreply@anthropic.com>
2025-07-10 08:41:34 +10:00

3.2 KiB
Raw Blame History

🐝 Hive MCP Server Service

This directory contains the systemd service configuration for running the Hive MCP Server as a background daemon with automatic agent discovery.

🚀 Quick Start

1. Install the Service

./install-service.sh

2. Start the Service

sudo systemctl start hive-mcp

3. Check Status

sudo systemctl status hive-mcp

4. View Logs

journalctl -u hive-mcp -f

🛠️ Management Script

Use the provided management script for easy operations:

# Install service
./hive-mcp.sh install

# Start/stop/restart
./hive-mcp.sh start
./hive-mcp.sh stop
./hive-mcp.sh restart

# Monitor
./hive-mcp.sh status
./hive-mcp.sh logs
./hive-mcp.sh follow

# Agent management
./hive-mcp.sh discover    # Trigger agent discovery
./hive-mcp.sh test       # Test backend connection

# Remove service
./hive-mcp.sh uninstall

⚙️ Configuration

The service is configured via environment variables in the service file:

  • HIVE_API_URL: Hive backend API endpoint (default: https://hive.home.deepblack.cloud/api)
  • HIVE_WS_URL: WebSocket endpoint (default: wss://hive.home.deepblack.cloud/socket.io)
  • AUTO_DISCOVERY: Enable periodic discovery (default: true)
  • DISCOVERY_INTERVAL: Discovery interval in ms (default: 300000 = 5 minutes)
  • LOG_LEVEL: Logging level (default: info)

🔄 Auto-Discovery

The service automatically:

  1. On Startup: Scans the network for available Ollama agents
  2. Periodically: Re-scans every 5 minutes (configurable)
  3. On Signal: Triggers discovery when receiving SIGHUP (systemctl reload hive-mcp)

📊 Monitoring

Service Status

sudo systemctl status hive-mcp

Live Logs

journalctl -u hive-mcp -f

Resource Usage

sudo systemctl show hive-mcp --property=MemoryCurrent,CPUUsageNSec

Agent Status

curl -s https://hive.home.deepblack.cloud/api/agents | jq

🔧 Troubleshooting

Service Won't Start

  1. Check logs: journalctl -u hive-mcp -n 50
  2. Verify backend connectivity: ./hive-mcp.sh test
  3. Check file permissions: ls -la /home/tony/AI/projects/hive/mcp-server/

Auto-Discovery Issues

  1. Check network connectivity to agent machines
  2. Verify Ollama is running on target machines
  3. Manually trigger discovery: ./hive-mcp.sh discover

High Resource Usage

  1. Check discovery interval: grep DISCOVERY_INTERVAL /etc/systemd/system/hive-mcp.service
  2. Monitor agent count: curl -s https://hive.home.deepblack.cloud/api/agents | jq '.total'
  3. Adjust memory limits in service file if needed

🛡️ Security

The service runs with:

  • Non-root user (tony)
  • Restricted filesystem access
  • Memory and CPU limits
  • Private tmp directory
  • No new privileges

📁 Files

  • hive-mcp.service - Systemd service definition
  • install-service.sh - Service installation script
  • hive-mcp.sh - Management script
  • logs/ - Log directory (created by service)
  • data/ - Data directory (created by service)

🔗 Integration

The service integrates with:

  • Hive Backend: https://hive.home.deepblack.cloud/api
  • Socket.IO: wss://hive.home.deepblack.cloud/socket.io
  • Systemd: Full systemd service lifecycle
  • Journal: Centralized logging via systemd-journald