Module 18 Quiz: Capstone Architecture

Instructions

Answer the following questions to test your knowledge of the architectural concepts covered in the capstone module.

Questions

1. What is the primary benefit of implementing “Golden Paths” in an Internal Developer Platform?

   • A) To force developers to use specific programming languages.
   • B) To reduce cognitive load by providing standardized, supported ways to deploy software.
   • C) To eliminate the need for CI/CD pipelines.
   • D) To increase the complexity of the deployment process.

2. When serving Large Language Models (LLMs) in production, what technique is used to maximize GPU utilization by processing multiple requests simultaneously?

   • A) Synchronous blocking
   • B) Continuous batching
   • C) Round-robin load balancing
   • D) Over-provisioning

3. Why is standard CPU-based autoscaling often insufficient for AI inference workloads?

   • A) Because AI models do not use the CPU.
   • B) Because GPU provisioning and loading large model weights take significant time, requiring specialized metrics like queue length.
   • C) Because Kubernetes cannot monitor CPU usage.
   • D) Because AI workloads are always perfectly predictable.

4. In a GitOps workflow (e.g., using ArgoCD), where is the “source of truth” for the desired state of the Kubernetes cluster?

   • A) The live cluster state
   • B) The developer’s local machine
   • C) The Git repository containing the declarative manifests
   • D) The container registry

5. Which observability pillar is most critical for debugging the latency of a single request as it travels through a microservices architecture and an AI inference engine?

   • A) Metrics
   • B) Logs
   • C) Distributed Tracing
   • D) Dashboards