MCP.so
登录
服务器

Multi-Cluster MCP server

@yanmxa

The gateway for GenAI systems to interact with open-cluster-management through the MCP

概览

What is Multi-Cluster MCP server?

Multi-Cluster MCP server is a Model Context Protocol (MCP) gateway that lets Generative AI systems interact with multiple Kubernetes clusters. It uses Open Cluster Management to perform operations on Kubernetes resources, manage multi-cluster environments, and deliver cluster observability.

How to use Multi-Cluster MCP server?

Configure the server in your MCP client using the JSON snippet with npx -y multicluster-mcp-server@latest. Ensure kubectl is installed and that the KUBECONFIG environment variable points to the hub cluster. The server accesses managed clusters through the hub cluster.

Key features of Multi-Cluster MCP server

  • Retrieve resources from the hub cluster (current context)
  • Retrieve resources from managed clusters
  • Connect to a managed cluster using a specified ClusterRole
  • Access resources across multiple Kubernetes clusters via Open Cluster Management
  • Retrieve and analyze metrics, logs, and alerts from integrated clusters
  • Interact with multi-cluster APIs, including Managed Clusters, Policies, and Add-ons

Use cases of Multi-Cluster MCP server

  • AI-driven troubleshooting across a fleet of Kubernetes clusters
  • Automated multi-cluster resource inspection and reporting
  • Centralized observability data retrieval for GenAI analysis
  • Policy and compliance checks across managed clusters

FAQ from Multi-Cluster MCP server

What prerequisites are needed to run the server?

You must have kubectl installed and the KUBECONFIG environment variable set to access the hub cluster.

How does the server access managed clusters?

It treats the configured cluster as the hub cluster and accesses managed clusters through it, optionally using a specified ClusterRole for connection.

Does the server support prompt templates or MCP resources?

Prompt templates and MCP resources for Open Cluster Management are currently in the planning stage and not yet available.

How do I configure the server in my MCP client?

Use the following JSON snippet:

{
  "mcpServers": {
    "multicluster-mcp-server": {
      "command": "npx",
      "args": ["-y", "multicluster-mcp-server@latest"]
    }
  }
}

What license is the server distributed under?

This project is licensed under the MIT License.

来自「其他」的更多内容