StarRocks Official MCP Server: Mirror & Scale

What is StarRocks Official MCP Server: Mirror & Scale?

The StarRocks MCP Server serves as a middleware layer connecting AI assistants with StarRocks databases. It enables seamless SQL execution and database exploration while eliminating the need for complex setup procedures. This server acts as both a mirror for database state and a scaling interface for managing database operations efficiently.

How to use StarRocks Official MCP Server: Mirror & Scale?

To configure the MCP server, start by defining environment variables specifying StarRocks host details. Two deployment methods are supported:

• Production environments: Use pre-packaged configurations

    {
      "mcpServers": {
        "mcp-server-starrocks": {
          "command": "uv",
          "args": ["run", "--with", "mcp-server-starrocks", "mcp-server-starrocks"],
          "env": {
            "STARROCKS_HOST": "default localhost",
            "STARROCKS_PORT": "default 9030",
            "STARROCKS_USER": "default root",
            "STARROCKS_PASSWORD": "default empty"
          }
        }
      }
    }
  

• Development environments: Reference local directories

    {
      "mcpServers": {
        "mcp-server-starrocks": {
          "command": "uv",
          "args": ["--directory", "path/to/mcp-server-starrocks", "run", "mcp-server-starrocks"],
          "env": {
            "STARROCKS_HOST": "default localhost",
            "STARROCKS_PORT": "default 9030",
            "STARROCKS_USER": "default root",
            "STARROCKS_PASSWORD": "default empty"
          }
        }
      }
    }
  

Key Features of StarRocks Official MCP Server: Mirror & Scale?

• Unified SQL Execution: Supports both read (SELECT) and write (DDL/DML) operations through dedicated tools
• Resource Navigation: Provides structured access to database objects via URI-based resource paths
• System Insights: Exposes internal StarRocks metadata through PROC filesystem emulation
• Deployment Flexibility: Adapts to both traditional and cloud-native StarRocks architectures

Use cases of StarRocks Official MCP Server: Mirror & Scale?

• Automated database management workflows
• Real-time monitoring of cluster health and performance
• Development environment setup acceleration
• Compliance auditing through transaction tracking
• Resource optimization using cluster balance analysis

FAQ from StarRocks Official MCP Server: Mirror & Scale?

• Q: What authentication methods are supported?
  A: Basic authentication via environment variables with optional password configuration
• Q: How are cloud-native deployments handled?
  A: Automatically detects compute nodes (CN) through PROC interfaces
• Q: Can custom SQL be executed?
  A: Yes, via the write_query tool for arbitrary commands
• Q: What happens if StarRocks is unreachable?
  A: The server returns connection status errors with diagnostic details

Starrocks Official MCP server

The StarRocks MCP Server acts as a bridge between AI assistants and StarRocks databases, allowing for direct SQL execution and database exploration without requiring complex setup or configuration.

Configuration

MCP server config

{
  "mcpServers": {
    "mcp-server-starrocks": {
      "command": "uv",
      "args": [
        "run",
        "--with",
        "mcp-server-starrocks",
        "mcp-server-starrocks"
      ],
      "env": {
        "STARROCKS_HOST": "default localhost",
        "STARROCKS_PORT": "default 9030",
        "STARROCKS_USER": "default root",
        "STARROCKS_PASSWORD": "default empty"
      }
    }
  }
}

If mcp-server-starrocks is not installed as python package(in dev env), can run using local dir

{
  "mcpServers": {
    "mcp-server-starrocks": {
      "command": "uv",
      "args": [
        "--directory",
        "path/to/mcp-server-starrocks",
        "run",
        "mcp-server-starrocks"
      ],
      "env": {
        "STARROCKS_HOST": "default localhost",
        "STARROCKS_PORT": "default 9030",
        "STARROCKS_USER": "default root",
        "STARROCKS_PASSWORD": "default empty"
      }
    }
  }
}

Components

Tools

• read_query

  • Execute a SELECT query or commands that return a ResultSet

• write_query

  • Execute an DDL/DML or other StarRocks command that do not have a ResultSet

Resources

Direct Resources

• starrocks:///databases
  • Lists all databases in StarRocks

Resource Templates

• starrocks:///{db}/{table}/schema

  • Gets the schema of a table using SHOW CREATE TABLE

• starrocks:///{db}/tables

  • Lists all tables in a specific database

• proc:///{+path}

  • System internal information exposed by StarRocks similar to linux /proc
  • Common paths include:
    • /frontends - Shows the information of FE nodes
    • /backends - Shows the information of BE nodes if this SR is non cloud native deployment
    • /compute_nodes - Shows the information of CN nodes if this SR is cloud native deployment
    • /dbs - Shows the information of databases
    • /dbs/<DB_ID> - Shows the information of a database by database ID
    • /dbs/<DB_ID>/<TABLE_ID> - Shows the information of tables by database ID
    • /dbs/<DB_ID>/<TABLE_ID>/partitions - Shows the information of partitions by database ID and table ID
    • /transactions - Shows the information of transactions by database
    • /transactions/<DB_ID> - Shows the information of transactions by database ID
    • /transactions/<DB_ID>/running - Shows the information of running transactions by database ID
    • /transactions/<DB_ID>/finished - Shows the information of finished transactions by database ID
    • /jobs - Shows the information of jobs
    • /statistic - Shows the statistics of each database
    • /tasks - Shows the total number of all generic tasks and the failed tasks
    • /cluster_balance - Shows the load balance information
    • /routine_loads - Shows the information of Routine Load
    • /colocation_group - Shows the information of Colocate Join groups
    • /catalog - Shows the information of catalogs

Prompts

None