Rtc MCP Server

🚀 Alibaba Cloud Realtime Compute MCP Server

An implementation of the Model Context Protocol (MCP) server for managing Apache Flink resources in Alibaba Cloud Realtime Compute. This server provides a standardized interface for AI models to interact with the Alibaba Cloud Flink service.

✨ Features

• Create and manage Flink clusters.
• Create and manage Flink SQL tasks.
• Deploy and control Flink applications.
• Monitor task status and metrics.
• Create and manage snapshot points.
• List and manage deployments.
• Manage workspaces and namespaces.

📦 Prerequisites

• JDK 17 or higher.
• Maven 3.6 or higher.
• An Alibaba Cloud account with access to Alibaba Cloud and permissions for the RTC Realtime Compute service.
• Alibaba Cloud access key ID and secret.

💻 Client Configuration

To use this server as an MCP client, add the following configuration to your MCP settings file (e.g., cline_mcp_settings.json):

{
  "mcpServers": {
    "rtc-mcp-server": {
      "command": "java",
      "args": [
        "-Dtransport.mode=stdio",
        "-Dspring.main.web-application-type=none",
        "-Dspring.main.banner-mode=off",
        "-Dlogging.file.name=/path/to/rtc-mcp-server/mcpserver.log",
        "-jar",
        "/path/to/rtc-mcp-server/target/rtc-mcp-server-1.0-SNAPSHOT.jar"
      ],
      "env": {
        "ALIYUN_ACCESS_KEY_ID": "your-access-key-id",
        "ALIYUN_ACCESS_KEY_SECRET": "your-access-key-secret"
      },
      "disabled": false,
      "autoApprove": []
    }
  }
}

Replace /path/to/rtc-mcp-server with the actual server path and provide your Alibaba Cloud credentials in the environment variables.

💻 Usage Examples

Available MCP Tools

This server provides the following MCP tools:

Task Management

• start_job: Start a deployed Flink task.
• stop_job: Stop a running Flink task.
• list_jobs: List all tasks in the deployment.
• delete_job: Delete a non - running task.
• get_job_diagnosis: Get task diagnosis information.

Deployment Management

• create_deployment: Create a new deployment.
• update_deployment: Update an existing deployment.
• delete_deployment: Delete a specified deployment.
• list_deployments: List all deployments.
• get_deployment_status: Get the status of a specific deployment.

Workspace Management

• create_workspace: Create a new workspace.
• update_workspace: Update the configuration of an existing workspace.
• delete_workspace: Delete a specified workspace.
• list_workspaces: List all workspaces.
• get_workspace_info: Get information about a specific workspace.

Resource Management

• allocate_resources: Apply for computing resources.
• release_resources: Release allocated resources.
• modify_resource_limits: Adjust resource limits.
• view_resource_usage: View resource usage.

🚀 Quick Start

Build the Project

mvn clean package

Run the Server

java -jar target/rtc-mcp-server-1.0-SNAPSHOT.jar

🔧 Technical Details

Server Modes

• Standard Mode: Provides full functionality and is suitable for production environments.
• Development Mode: Enhances debugging capabilities and is suitable for development and testing phases.

Logging

By default, logs are output to the mcpserver.log file. You can adjust the configuration as follows:

logging.file.name=/path/to/log/mcpserver.log
logging.level=INFO

🤝 Contribution Guidelines

Community contributions are welcome! Please refer to the project documentation to learn how to participate in development and submit code.

📄 License

This project is licensed under the Apache License 2.0. The specific content is as follows:

Unless required by applicable law or agreed to in writing, software and its documentation under the above license can be used freely in any case.