GreenDoc - SUSEAI Demo Lab

What is GreenDoc - SUSEAI Demo Lab?

GreenDoc - SUSEAI Demo Lab is a document that describes how to implement a local lab to demonstrate capabilities of SUSE AI in the context of a specific use case. It provides step-by-step instructions to set up a minimal SUSE AI stack on a K3s cluster using Rancher Fleet and the SUSE Application Collection, including an MCP server and Open WebUI.

How to use GreenDoc - SUSEAI Demo Lab?

Follow the cluster setup: create namespaces and secrets for the SUSE Application Collection, deploy cert-manager and Rancher Fleet via Helm. Then apply a GitRepo configuration (suseai.yaml) to deploy the SUSE AI stack with a Kubernetes MCP server. Finally, configure Open WebUI’s “Tools Servers” (e.g., http://mcp-service.suseai.svc.cluster.local:8000/kubernetes) and enable the tool for a model.

Key features of GreenDoc - SUSEAI Demo Lab

• Uses K3s as the lightweight Kubernetes distribution.
• Automates deployment with Rancher Fleet.
• Leverages the SUSE Application Collection (AppCo) for components.
• Includes a Kubernetes MCP server (from manusa/kubernetes-mcp-server).
• Integrates with Open WebUI for AI assistant interactions.
• Provides a ready-to-demo environment with minimal configuration.

Use cases of GreenDoc - SUSEAI Demo Lab

• Demonstrating SUSE AI capabilities in a local lab.
• Showing Kubernetes management via an MCP server from an AI chatbot.
• Evaluating Open WebUI tool integration with MCP servers.
• Creating a reproducible demo environment for SUSE AI workflows.

FAQ from GreenDoc - SUSEAI Demo Lab

What prerequisites are needed to set up this demo lab?

You need a Kubernetes cluster (K3s is used in the guide), credentials for the SUSE Application Collection (username and token), and Helm installed to deploy cert-manager and Rancher Fleet.

How do I access the Kubernetes MCP server after deployment?

Open a browser and go to http://localhost:8000. Click on the kubernetes link. Authorize using the token top-secret and then use the "try out" option to test queries.

How is the MCP server connected to Open WebUI?

In Open WebUI admin settings, under Tools, add a new tool server with the URL http://mcp-service.suseai.svc.cluster.local:8000/kubernetes. Then enable "Function Calling" to "native" and select the tool for the model you want to use.

What data or resources does this demo use?

All components are deployed inside the local K3s cluster. Secrets are created from the SUSE Application Collection token. No external data is pulled beyond the Helm charts and container images from the Application Collection.

What transport and authentication does the MCP server use?

The MCP server uses HTTP (port 8000). Authentication is token-based; the example token is top-secret (configured via the authorize button in the web UI).