GCP Associate Cloud Engineer Practice Question

Your team has packaged a stateless Go web API into a container image. Traffic is unpredictable: some days the service is idle, while marketing campaigns can generate tens of thousands of simultaneous HTTP requests within minutes. Operations wants the platform to scale to zero when idle, scale up automatically without managing nodes or clusters, and support setting per-request concurrency limits in the future. Which Google Cloud compute option best meets these requirements?

Deploy the container to a GKE Autopilot cluster with Horizontal Pod Autoscaling.
Create an autoscaled managed instance group with preemptible VMs on Compute Engine.
Deploy the container to Cloud Run.
Rewrite the service as Cloud Functions triggered by HTTP.

Report Issue

Answer Description

Cloud Run is designed for containerized, stateless services that respond to HTTP requests. It is fully managed, so there is no need to provision or manage servers or Kubernetes clusters. Cloud Run automatically scales instances from zero to thousands based on incoming traffic and allows you to configure the maximum number of concurrent requests each instance can handle.

GKE Autopilot removes node management but never scales to zero; you continue paying for the control plane and a minimum set of nodes even when no traffic arrives. A managed instance group on Compute Engine also keeps at least one VM running and requires OS maintenance. Cloud Functions provides scale-to-zero, but first-generation functions handle exactly one request at a time and are optimized for small event-driven functions rather than full web APIs packaged as arbitrary containers, making it less flexible than Cloud Run for this use case.

Ask Bash

Bash is our AI bot, trained to help you pass your exam. AI Generated Content may display inaccurate information, always double-check anything important.