noc-mcp-server

venkatbongoni/noc-mcp-server

3.1

If you are the rightful owner of noc-mcp-server and would like to certify it and/or have it hosted online, please leave a comment on the right or send an email to dayong@mcphub.com.

NOC Internal MCP server is designed to manage and streamline internal tools such as syslog, Splunk, command execution, memory management, and knowledge sharing.

Tools
5
Resources
0
Prompts
0

NOC Internal MCP Server for All Internal Tools

A comprehensive Model Context Protocol (MCP) server implementation with DSPY framework integration for Network Operations Center (NOC) tools including syslog analysis, Splunk integration, command execution, memory management, and knowledge base operations.

🎯 Overview

This project builds an MCP server that exposes network operations tools through a standardized protocol, enabling seamless integration with DSPY agents and other MCP-compatible clients. The server provides both simplified demonstration capabilities and full production-ready functionality.

🏗️ Architecture

Core Components

noc-mcp-server/
├── core/                    # Core MCP server functionality
│   ├── mcp_server.py       # Main MCP server implementation
│   ├── mcp_client.py       # MCP client for connecting to servers
│   ├── mcp_agent.py        # DSPY agent with MCP integration
│   ├── mcp_tools.py        # Full MCP tool wrappers
│   ├── mcp_tools_simple.py # Simplified tool wrappers for testing
│   └── tool_registry.py    # Tool package management
├── agents/                  # Agent implementations
│   └── syslog_tools.py     # Original syslog tool definitions
├── auth/                    # Authentication management
│   └── auth_manager.py     # Authentication handling
├── config/                  # Configuration files
│   └── tool_packages.yaml  # Tool package definitions
├── tools/                   # Original tool implementations (DO NOT MODIFY)
├── lib/                     # Shared libraries
│   ├── datastore.py        # Memory management
│   └── datastore_interface.py
├── tests/                   # Test suite
└── scripts/                 # Utility scripts

Tool Packages

The server supports different tool packages defined in config/tool_packages.yaml:

  • syslog: System log analysis tools

    • READ_NEW_LOG_MESSAGES: Read recent syslog data
    • HIGHEST_SERVICE_IMPACT: Evaluate operational impact
    • QUERY_DEVICES: Execute commands on network devices
    • HOLISTIC_CAUSAL_REASONING: Root cause analysis
    • EVALUATE_DRAFT_ANSWER: Analysis completeness evaluation
    • SHORT_TERM_MEMORY: Memory operations

  • rag: Retrieval-Augmented Generation tools

    • LOOK_UP_FROM_ICS_SQL_SYSTEM: ICS SQL system lookup
    • LOOK_UP_FROM_ICS_RAGSYSTEM: ICS RAG system lookup

🚀 Quick Start

Prerequisites

  • Python >3.13, <3.14
  • Virtual environment (recommended)

Installation

  1. Clone and setup the environment:

    cd noc-mcp-server
chmod +x setup.sh
./setup.sh

  2. Activate the virtual environment:

    source venv/bin/activate

  3. Run basic functionality test:

    python3 test_basic.py

Running the Server

Option 1: Simplified Server (Recommended for Testing)

Run the simplified server with automatic demo:

python3 simple_server.py --auto-demo

Run interactive simplified server:

python3 simple_server.py

Available commands in interactive mode:

  • workflow <query> - Run full analysis workflow
  • call <tool> <key=value> - Call specific tool
  • memory list - List memory contents
  • memory get <key> - Get memory value
  • help - Show help
  • quit - Exit
Option 2: Full MCP Server (Requires Dependencies)
python -m core.mcp_server --port 8075 --tool-package syslog

Example Usage

Simplified Server Demo
$ python3 simple_server.py --auto-demo

============================================================
AUTO DEMO: High CPU utilization detected
============================================================

READ_LOGS:
  Retrieved 3 log messages for query: High CPU utilization detected

IMPACT_ANALYSIS:
  Impact analysis for 'Impact analysis for: High CPU utilization detected': Found 1 high-impact events out of 3 total events

DEVICE_QUERY:
  Device query for 'Device diagnostics for: High CPU utilization detected': Checked 3 devices
  router1: up
  switch1: down
  router2: up

ROOT_CAUSE:
  Root cause analysis for 'Root cause for: High CPU utilization detected':
  • High impact events detected: 1
  • Device failures detected: switch1
  • Link state changes detected: 1 events

Recommendation: Investigate device failures and link state changes

EVALUATION:
  Draft answer evaluation for 'Evaluate completeness for: High CPU utilization detected':
Completeness Score: 100/100
...
Interactive Demo
$ python3 simple_server.py

noc> workflow Network latency issues
Running workflow: Network latency issues
...

noc> call SHORT_TERM_MEMORY action=list
Memory: Memory keys: context, high_impact, device_data

noc> memory get context
Memory: ['2024-01-15 10:00:01 router1: %BGP-5-ADJCHANGE...', ...]

🔧 Configuration

Environment Variables

Create a .env file (automatically created by setup script):

# Server settings
NOC_MCP_PORT=8075
NOC_TOOL_PACKAGE=syslog

# Logging
LOG_LEVEL=INFO

# Syslog configuration
SYSLOG_API_BASE=http://localhost:8080
SYSLOG_HOST=localhost
SYSLOG_BUFFER=main

# OpenAI Configuration (optional)
OPENAI_API_KEY=your_api_key_here
MODEL_NAME=gpt-4o-mini

Tool Package Configuration

Edit config/tool_packages.yaml to define custom tool packages:

syslog:
  - READ_NEW_LOG_MESSAGES
  - HIGHEST_SERVICE_IMPACT
  - QUERY_DEVICES
  - HOLISTIC_CAUSAL_REASONING
  - EVALUATE_DRAFT_ANSWER
  - SHORT_TERM_MEMORY

custom_package:
  - CUSTOM_TOOL_1
  - CUSTOM_TOOL_2

🧪 Testing

Basic Functionality Test

python3 test_basic.py

Unit Tests

python -m pytest tests/ -v

Integration Tests

python integrated_demo.py --duration 30

🏭 Production Deployment

Full Dependencies Installation

For production deployment with all features:

pip install -r requirements.txt
pip install fastmcp mcp dspy-ai

Running with Full MCP Support

python -m core.mcp_server --port 8075 --tool-package syslog --log-level INFO

📊 Monitoring and Observability

Logging

The server provides comprehensive logging:

  • Request/response logging
  • Tool execution tracking
  • Error monitoring
  • Performance metrics

Health Checks

# Via simplified server
noc> call health_check

# Via MCP client
curl http://localhost:8075/health

🔌 Integration Examples

DSPY Agent Integration

from core.mcp_agent import NOCMCPAgent

async def analyze_network_issue():
    async with NOCMCPAgent("http://localhost:8075") as agent:
        result = await agent.analyze_network_issue("BGP neighbor down")
        print(result["final_analysis"])

MCP Client Integration

from core.mcp_client import NOCMCPClient

async def use_tools():
    async with NOCMCPClient("http://localhost:8075") as client:
        tools = client.get_available_tools()
        result = await client.call_tool("READ_NEW_LOG_MESSAGES", 
                                      query="Recent network errors")

🛠️ Development

Adding New Tools

  1. Create tool wrapper in core/mcp_tools_simple.py:
class NewToolWrapper(MCPToolWrapper):
    def __init__(self):
        super().__init__("NEW_TOOL", "Description of new tool")
    
    async def execute(self, **kwargs) -> str:
        # Tool implementation
        return "Tool result"
  1. Register in factory method:
tool_map = {
    "NEW_TOOL": NewToolWrapper,
    # ... existing tools
}
  1. Add to tool package in config/tool_packages.yaml:
syslog:
  - NEW_TOOL
  # ... existing tools

🚨 Troubleshooting

Common Issues

  1. Import Errors: Ensure virtual environment is activated and dependencies installed
  2. Port Conflicts: Change port with --port argument
  3. Tool Not Found: Check tool name in config/tool_packages.yaml
  4. Memory Issues: Use memory clear or restart server

Debug Mode

python3 simple_server.py --log-level DEBUG

📚 API Reference

Tool Signatures

All tools follow the MCP tool signature pattern:

async def tool_name(**kwargs) -> str:
    """Tool description.
    
    Args:
        query: Analysis query (optional)
        **kwargs: Tool-specific parameters
        
    Returns:
        str: Tool execution result
    """

Memory Operations

# Set memory
await call_tool("SHORT_TERM_MEMORY", action="set", key="data", value="value")

# Get memory  
await call_tool("SHORT_TERM_MEMORY", action="get", key="data")

# List keys
await call_tool("SHORT_TERM_MEMORY", action="list")

# Clear all
await call_tool("SHORT_TERM_MEMORY", action="clear")

🤝 Contributing

  1. Fork the repository
  2. Create feature branch: git checkout -b feature/new-feature
  3. Make changes and test: python3 test_basic.py
  4. Commit changes: git commit -am 'Add new feature'
  5. Push to branch: git push origin feature/new-feature
  6. Submit pull request

📄 License

MIT License - see LICENSE file for details.

🆘 Support

For issues and questions:

  1. Check troubleshooting section
  2. Run basic tests: python3 test_basic.py
  3. Check logs for error details
  4. Create GitHub issue with detailed description

Status: ✅ Basic functionality implemented and tested Next Steps: Full MCP server implementation with external dependencies