Capstone Lab: Building the Foundation
Overview
In this lab, you will bootstrap the core infrastructure required for your final capstone project. You will use Terraform to provision a Kubernetes cluster and deploy the foundational components of both an Internal Developer Platform and an AI Inference Engine.
Scenario
You are the lead platform engineer for a fast-growing startup. Your task is to build a platform that allows developers to easily deploy AI-powered applications.
Steps
1. Provision the Kubernetes Cluster
Use the provided Terraform modules to spin up an EKS/GKE cluster with a dedicated GPU node pool.
2. Install the IDP Core
Deploy Backstage (or your chosen IDP) to the cluster. Configure it to connect to your Git provider for software catalog discovery.
3. Deploy the AI Inference Engine
Deploy vLLM to the GPU node pool. Verify that it can successfully load a small, open-source model (e.g., Llama-3-8B).
4. Configure GitOps
Install ArgoCD and configure it to manage the state of the IDP and AI components.
Validation
Your lab is complete when:
- The developer portal is accessible via an Ingress.
- You can successfully query the vLLM endpoint and receive a response.
- All components are managed by ArgoCD and synced with your Git repository.