Skip to main content
Glama

drain-node

Drain a Kubernetes node for maintenance by safely evicting pods, with options to force drain or ignore DaemonSets.

Instructions

Drain a node for maintenance

Input Schema

TableJSON Schema
NameRequiredDescriptionDefault
nodeYesThe name of the node to drain
forceNoForce drain even if there are unmanaged pods
ignore_daemonsetsNoIgnore DaemonSet pods during drain

Implementation Reference

  • Executes the drain-node tool by running the 'kubectl drain' command with optional force and ignore-daemonsets flags, returning the command output.
    case "drain-node": {
      const { node, force = false, ignore_daemonsets = true } = args;
      const forceArg = force ? "--force" : "";
      const ignoreArg = ignore_daemonsets ? "--ignore-daemonsets" : "";
      const cmd = `kubectl drain ${node} ${forceArg} ${ignoreArg} --delete-emptydir-data`;
      const { stdout } = await execAsync(cmd);
      return {
        content: [{ type: "text", text: stdout || `Node ${node} drained successfully` }]
      };
    }
  • Tool definition including name, description, and input schema for the drain-node tool, used for registration and validation.
      name: "drain-node",
      description: "Drain a node for maintenance",
      inputSchema: {
        type: "object",
        properties: {
          node: { 
            type: "string",
            description: "The name of the node to drain"
          },
          force: { 
            type: "boolean",
            description: "Force drain even if there are unmanaged pods"
          },
          ignore_daemonsets: { 
            type: "boolean",
            description: "Ignore DaemonSet pods during drain"
          }
        },
        required: ["node"]
      }
    },
  • server.js:1392-1394 (registration)
    Registers the list of all tools, including drain-node, for the ListToolsRequest.
    server.setRequestHandler(ListToolsRequestSchema, async () => {
      return { tools };
    });

Latest Blog Posts

MCP directory API

We provide all the information about MCP servers via our MCP API.

curl -X GET 'https://glama.ai/api/mcp/v1/servers/thekaranpargaie/kube-mcp'

If you have feedback or need assistance with the MCP directory API, please join our Discord server