RDSMCP

bobtherdsman/RDSMCP

3.1

If you are the rightful owner of RDSMCP 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.

The RDS MCP - SQL Server to RDS Migration Assessment Tool is a Python-based utility designed to evaluate SQL Server instances for migration to AWS RDS, offering both CLI and MCP server modes for integration with AI assistants.

Tools
3
Resources
0
Prompts
0

RDS MCP - SQL Server to RDS Migration Assessment Tool

A Python-based tool for assessing SQL Server instances for migration to AWS RDS. Functionally equivalent to PowerShell RDSDiscoveryGuide for automated technical discovery. Available as both a CLI tool and an MCP (Model Context Protocol) server for integration with AI assistants.

Features

  • 100% PowerShell Compatibility: Identical SQL queries and RDS compatibility logic
  • Comprehensive Assessment: Analyzes 24 SQL Server features for RDS compatibility
  • Dual Modes: Use as standalone CLI or MCP server
  • Batch Processing: Analyze multiple servers from a file
  • Instance Recommendations: Get RDS instance type recommendations (db.m* family, no t3)
  • CSV Export: 39-column format matching PowerShell technical output
  • DBC Export: 23-column Database Consolidation format for migration planning
  • SSIS/SSRS Detection: Identifies Integration/Reporting Services (informational only)
  • Enterprise Features: Detects Enterprise-only features without blocking RDS migration
  • Automated Technical Focus: Excludes business survey questions

Installation

Prerequisites

Python:

  • Python 3.8 or higher
  • pip package manager

ODBC Driver for SQL Server:

The tool requires Microsoft ODBC Driver 18 for SQL Server.

Linux (Ubuntu/Debian):

curl https://packages.microsoft.com/keys/microsoft.asc | sudo apt-key add -
curl https://packages.microsoft.com/config/ubuntu/$(lsb_release -rs)/prod.list | sudo tee /etc/apt/sources.list.d/mssql-release.list
sudo apt-get update
sudo ACCEPT_EULA=Y apt-get install -y msodbcsql18

Linux (RHEL/CentOS):

curl https://packages.microsoft.com/config/rhel/8/prod.repo | sudo tee /etc/yum.repos.d/mssql-release.repo
sudo yum remove unixODBC-utf16 unixODBC-utf16-devel
sudo ACCEPT_EULA=Y yum install -y msodbcsql18

macOS:

brew tap microsoft/mssql-release https://github.com/Microsoft/homebrew-mssql-release
brew update
HOMEBREW_ACCEPT_EULA=Y brew install msodbcsql18

Windows:

Install RDS MCP Tool

# Clone repository
git clone <repo-url>
cd rdsmcp

# Create virtual environment (recommended)
python3 -m venv venv
source venv/bin/activate  # On Windows: venv\Scripts\activate

# Install dependencies
pip install -r requirements.txt

Verify Installation

# Test CLI
python3 cli.py --help

# Test connection to SQL Server
python3 cli.py analyze --host <your-server> --username <user> --password <password>

Usage

CLI Mode

Single Server Analysis

SQL Authentication:

python cli.py analyze --host <hostname> --username <user> --password <pass> [--port 1433]

Windows Authentication:

python cli.py analyze --host <hostname> --windows-auth [--port 1433]
Batch Analysis

SQL Authentication:

python cli.py batch --input servers.txt --username <user> --password <pass> --output results.csv

Windows Authentication:

python cli.py batch --input servers.txt --windows-auth --output results.csv

With DBC.CSV output:

python cli.py batch --input servers.txt --username <user> --password <pass> --output results.csv --dbc

This generates both results.csv (39 columns) and results_DBC.csv (23 columns)

Input file format (one server per line):

server1.example.com
server2.example.com
10.0.1.100
Instance Recommendation
python cli.py recommend --cpu 16 --memory 64 --edition SE --version 15

MCP Server Mode

Configure in your MCP client (e.g., Kiro CLI):

{
  "mcpServers": {
    "rds-discovery": {
      "command": "/path/to/venv/bin/python",
      "args": ["/path/to/rdsmcp/server.py"]
    }
  }
}

Available MCP tools:

  • analyze_sql_server - Analyze single SQL Server instance
  • recommend_rds_instance - Get RDS instance recommendations
  • analyze_sql_servers_batch - Batch analyze multiple servers with CSV export

Assessment Checks

The tool evaluates the following features for RDS compatibility (matches PowerShell logic exactly):

Blocking Features (RDS Incompatible)

  1. Database Count > 100
  2. Linked Servers (heterogeneous - non-SQL Server/Oracle)
  3. Log Shipping
  4. FILESTREAM
  5. Resource Governor
  6. Transaction Replication
  7. Extended Procedures (non-standard)
  8. TSQL Endpoints
  9. PolyBase
  10. File Tables
  11. Buffer Pool Extension
  12. Stretch Database
  13. Trustworthy Databases
  14. Server Triggers
  15. Machine Learning Services (R/Python)
  16. Policy-Based Management
  17. Data Quality Services
  18. CLR Enabled (SQL 2017+)

Informational Only (Not Blocking)

  • Always On Availability Groups - Current state, not a blocker
  • Always On Failover Cluster Instances - Current state, not a blocker
  • Server Role (Primary/Secondary/Readable/Standalone) - Informational
  • Enterprise Edition Features (Partitioning, Compression, CDC, etc.) - Supported in RDS EE
  • SSIS (SQL Server Integration Services) - Separate service, migrate to AWS Glue/EC2
  • SSRS (SQL Server Reporting Services) - Separate service, migrate to EC2/QuickSight
  • Read Only Replica - Configuration detail

RDS Custom Compatibility

  • RDS Standard: Up to 16 TB database size
  • RDS Custom: Up to 16 TB (same limit)
  • Above 16 TB: EC2 recommended

Output Format

Standard CSV (39 columns)

Full technical assessment with all feature checks, compatibility status, and recommendations.

DBC CSV (23 columns)

Simplified Database Consolidation format for migration planning:

  • ServerName, VCPU, Memory, Edition
  • Cluster/Always On status
  • Instance recommendations
  • Placeholders for manual input (CPU/Memory utilization, storage, IOPS, throughput)
  • Optimized for infrastructure planning and capacity estimation

Generate with --dbc flag: python cli.py batch --input servers.txt --username user --password pass --dbc

CSV Columns (39 total)

  1. Server Name
  2. SQL Server Current Edition
  3. SQL Server current Version
  4. Sql server Source (RDS/EC2/OnPrem)
  5. SQL Server Replication
  6. Heterogeneous linked server
  7. Database Log Shipping
  8. FILESTREAM
  9. Resource Governor
  10. Service Broker Endpoints
  11. Non Standard Extended Proc
  12. TSQL Endpoints
  13. PolyBase
  14. File Table
  15. buffer Pool Extension
  16. Stretch DB
  17. Trust Worthy On
  18. Server Side Trigger
  19. R & Machine Learning
  20. Data Quality Services
  21. Policy Based Management
  22. CLR Enabled
  23. DB count Over 100
  24. Total DB Size in GB
  25. Always ON AG enabled
  26. Always ON FCI enabled
  27. Read Only Replica
  28. Server Role Desc
  29. RDS Compatible
  30. RDS Custom Compatible
  31. EC2 Compatible
  32. Elasticache
  33. Enterprise Level Feature Used
  34. Memory
  35. CPU
  36. Instance Type
  37. isSSIS
  38. isSSRS
  39. Note

JSON Output (MCP/CLI)

{
  "server_info": {
    "edition": "Standard Edition (64-bit)",
    "version": "16.0.4095.4",
    "is_clustered": false,
    "source": "EC2/OnPrem"
  },
  "resources": {
    "cpu": 16,
    "max_memory_mb": 32768,
    "total_db_size_gb": 100.5
  },
  "features": {
    "linked_servers": "N",
    "filestream": "N",
    "server_triggers": "N",
    "always_on_ag": "N",
    "enterprise_features": "",
    "ssis": "N",
    "ssrs": "N"
  },
  "rds_compatible": true,
  "recommendation": {
    "primary_recommendation": "db.m6i.4xlarge",
    "recommended_instances": ["db.m6i.4xlarge", "db.m5d.4xlarge"],
    "type": "G"
  }
}

Instance Recommendations

  • General Purpose (G): db.m5d., db.m6i. (no t3 burstable instances)
  • Memory Optimized (M): db.r*, db.x*, db.z1d.* (excludes db.m*, db.r3*, db.r4*, db.t3*, db.x1*, db.x1e*)
  • High Memory (>1TB): db.x* family
  • Sizing: Based on CPU/4 ratio, matches PowerShell logic

Requirements

  • Python 3.8+
  • pyodbc
  • ODBC Driver 18 for SQL Server
  • SQL Server credentials with appropriate permissions OR Windows Authentication
  • AwsInstancescsv.csv (RDS instance data)

Authentication Methods

SQL Authentication

  • Requires username and password
  • Works on Windows and Linux
  • Example: --username sa --password MyPassword123

Windows Authentication (Trusted Connection)

  • Uses current Windows user credentials
  • Requires --windows-auth flag
  • No username/password needed
  • Windows only (or Linux with Kerberos configured)
  • Example: --windows-auth

Permissions Required

The tool requires the following SQL Server permissions:

  • VIEW SERVER STATE
  • VIEW ANY DEFINITION
  • Access to system databases (master, msdb)
  • sp_MSforeachdb execution (for Enterprise Features, File Tables, SSIS detection)

Differences from PowerShell

Not Included (Intentional)

  • Business Questions (L100): 9 manual input questions excluded (requires human input)
  • Elasticache Analysis: Read/write pattern analysis not implemented
  • Utilization-Based Scaling: CPU/Memory utilization adjustments not implemented

Improvements Over PowerShell

  • Windows Authentication Support: Added Trusted Connection support (PowerShell has this)
  • More Accurate RDS Compatibility: Correctly excludes SSIS/SSRS/Enterprise Features/Always On from blocking
  • Correct RDS Custom Limit: 16TB (PowerShell uses 14.5TB)
  • SSIS/SSRS Columns: Added to CSV output (PowerShell doesn't include)
  • Dynamic Notes: Adds SSIS/SSRS detection info to notes field
  • Better Error Handling: Detailed error capture with summary statistics
  • JSON Output: Available in addition to CSV
  • MCP Integration: Can be used as tool by AI assistants

License

MIT License

Support

For issues or questions, please open an issue on GitHub.