erikdarlingdata/PerformanceMonitor

Free, open-source SQL Server and Postgres performance monitoring. Collectors, real-time alerts, graphical plan viewer, MCP server for AI analysis. Supports SQL 2016-2025, Azure SQL, AWS RDS, Postgres, Aurora Postgres.

What it solves

PerformanceMonitor is a free, open-source alternative to expensive proprietary SQL Server monitoring tools. It provides real-time visibility into server health, query performance, and system bottlenecks without requiring expensive licenses or complex server-side installations.

How it works

The project offers two primary editions based on the user's needs:

  • Lite Edition: A standalone desktop application that monitors servers remotely on-demand. It uses a local DuckDB and Parquet files for data storage, ensuring no installation is required on the target SQL Server.
  • Darling Edition: A headless Windows service designed for 24/7 always-on monitoring of multiple servers. It collects data into a central PostgreSQL store (with automatic TimescaleDB adoption for compression) and provides a detached viewer and an optional web dashboard.

Both editions utilize a shared library of 42 specialized collectors that gather data on wait stats, blocking chains, deadlock graphs, and memory grants. It also includes a built-in MCP server that allows AI tools like Claude to analyze the monitoring data directly.

Who it’s for

Database Administrators (DBAs), consultants, and IT operators who need professional-grade SQL Server monitoring, real-time alerting, and performance tuning tools without the cost of commercial software.

Highlights

  • Zero Server Footprint: The Lite edition requires no installation on the target server, making it ideal for locked-down environments.
  • AI-Powered Analysis: Includes a built-in MCP server that enables AI agents to query and analyze server performance data.
  • Comprehensive Monitoring: 42 specialized collectors covering everything from CPU utilization and file I/O to deadlock graphs and execution plans.
  • Integrated Tuning Tools: Features a graphical plan viewer with a 30-rule PlanAnalyzer and a recommendations engine that can apply fixes directly.
  • Privacy-First: No telemetry or cloud dependencies; all data remains on the user's local machine or server.

Related

  • Project
    wshobson/maverick-mcp

    A personal stock analysis MCP server that provides financial data, technical indicators, and portfolio tracking tools to AI assistants.

  • Project
    bvisible/mcp-ssh-manager

    An MCP server that enables AI agents to securely manage multiple SSH remote servers, execute commands, and automate DevOps tasks with per-server security restrictions.

  • Project
    TabularisDB/tabularis

    An open-source desktop SQL workspace for multiple databases featuring SQL notebooks, a built-in MCP server for AI agents, and local AI text-to-SQL capabilities.

  • Project
    googleapis/mcp-toolbox

    An open-source MCP server that connects AI agents and IDEs to enterprise databases through prebuilt generic tools or a customizable framework for secure, structured data access.

  • Project
    ClickHouse/mcp-clickhouse

    An MCP server that enables AI agents to interact with ClickHouse databases and the embedded chDB engine to execute SQL queries and explore data.