Examkin

Soru

Zorluk: ZorSelecting and Designing Compute and Application Platforms

A digital publishing firm is architecting a new public-facing event notification backend on Google Cloud. The system processes incoming HTTPS webhooks triggered by third-party content providers. The workload is strictly stateless, experiences extreme traffic variance ranging from extended periods of zero traffic to sudden bursts of thousands of requests per second, and requires container-level custom runtime dependencies. The architecture mandate prioritizes minimizing operational overhead, eliminating baseline infrastructure idle costs, and avoiding container cluster management. Which compute platform selection best fulfills these technical and business requirements?

  1. Deploy the containerized application on Cloud Run.Cevap
  2. B
    Deploy the containerized application on a GKE Autopilot cluster using Horizontal Pod Autoscaler (HPA).
  3. C
    Deploy the application across a Managed Instance Group (MIG) of Compute Engine Virtual Machines configured with CPU-based autoscaling.
  4. D
    Deploy the application on App Engine Flexible Environment with automatic scaling.

Cevap

Deploying the containerized application on Cloud Run is the optimal architecture choice.
Cloud Run allows executing stateless containerized workloads over HTTP/HTTPS with automatic scaling down to zero when idle. This fulfills all constraints by eliminating baseline idle costs, accommodating unpredictable traffic spikes rapidly, and completely removing cluster and infrastructure management overhead.

Adım Adım Çözüm

1
Analyze workload state and protocol requirements.
The application is containerized, stateless, and handles incoming HTTPS webhooks.
Containerized HTTP workloads are well-suited for serverless container platforms or Kubernetes.
2
Evaluate operational overhead and scaling constraints.
Requirements mandate zero cluster management, rapid response to high traffic variance, and scale-to-zero capability.
Scale-to-zero eliminates idle billing during zero-traffic periods, while managed serverless eliminates node/cluster lifecycle maintenance.
3
Select the matching GCP compute platform.
Cloud Run provides serverless container execution with automatic scaling to zero, minimal operational overhead, and per-request billing.
GKE introduces unnecessary cluster abstraction (overkill), Compute Engine MIGs incur VM boot times and baseline VM costs, and App Engine Flexible cannot scale to zero instances.

Anahtar Kavram

Compute Platform Selection for Serverless Containers
Bu soruyu puanla

Konular

Selecting and Designing Compute and Application Platforms