Question 8
Domain 5: Deployment, Scaling, Safety, and ComplianceYou are deploying a multi-agent AI platform using Docker containers orchestrated by Kubernetes. The system must autoscale based on traffic while ensuring requests are evenly distributed across all agent replicas. Which combination of tools and configurations best satisfies these requirements?
Correct answer: C
Explanation
Kubernetes provides orchestration for containerized agents, while HPA “autoscale[s] based on traffic” by adjusting replica counts as demand changes. A Service of type LoadBalancer, often paired with Ingress, exposes the application and distributes incoming requests across pods so traffic is evenly routed among agent replicas.
Why each option is right or wrong
A. Docker Compose deployment using NodePort services for inter-container traffic
B. Kubernetes with ClusterIP services and a Horizontal Pod Autoscaler scaling on CPU
C. Kubernetes with Ingress, HPA, and service type LoadBalancer
Kubernetes is the correct orchestration layer for Dockerized agents because it manages pod scheduling and replica sets, and the Horizontal Pod Autoscaler (HPA) is the built-in controller that changes replica counts automatically based on observed metrics such as CPU or custom traffic signals, with scaling behavior governed by the HPA API in Kubernetes. Exposing the workload through a Service of type LoadBalancer creates an external entry point that forwards traffic to the backing pods, while Ingress provides HTTP/HTTPS routing to the service so requests are distributed across the available replicas rather than pinned to a single container.
D. Docker Swarm with host networking on each node for low-latency intra-cluster traffic