Hard1 markMultiple Choice
Domain 5: Managing Implementation and Ensuring ReliabilityDomain 5GKE AutoscalingCloud CDNPerformance Optimization
This question is part of a case study — click to read the full scenario(Case 06)

CASE STUDY: TrendWear Apparel

Company Overview:
TrendWear Apparel is a global clothing retailer with an e-commerce platform and 500 physical stores.

Current Technical Environment:

  • On-premises VMware environment
  • Legacy IBM Mainframe for core inventory management
  • Monolithic e-commerce application running on VMs

Business Requirements:

  • Modernize the e-commerce platform to handle Black Friday (10x normal traffic)
  • Unify online and in-store inventory data in real-time
  • Avoid major capital expenditure (CapEx) for data center refreshes

Executive Statements:

  • CEO: "We need an omnichannel experience. Customers should see accurate store inventory online."
  • CFO: "We must shift from CapEx to OpEx. No more buying hardware."
  • CTO: "We want to move to microservices, but we cannot retire the mainframe for at least 3 years due to complex legacy dependencies."

Technical Requirements:

  • Hybrid architecture connecting GCP and on-premises
  • Microservices architecture for the new e-commerce platform
  • PCI-DSS compliance for all payment processing
  • Consistent management plane across on-prem and cloud

Constraints:

  • Mainframe must remain on-premises
  • E-commerce migration must be completed before the next holiday season (8 months)

QUESTION:
Based on the CTO's requirement for a consistent management plane across on-premises and the cloud, and the need to modernize to microservices while keeping the mainframe, which compute architecture should you recommend?

GCP PCA · Question 10 · Domain 5: Managing Implementation and Ensuring Reliability

CASE STUDY: TrendWear Apparel

Company Overview:
TrendWear Apparel is a global clothing retailer with an e-commerce platform and 500 physical stores.

Current Technical Environment:

  • On-premises VMware environment
  • Legacy IBM Mainframe for core inventory management
  • Monolithic e-commerce application running on VMs

Business Requirements:

  • Modernize the e-commerce platform to handle Black Friday (10x normal traffic)
  • Unify online and in-store inventory data in real-time
  • Avoid major capital expenditure (CapEx) for data center refreshes

Executive Statements:

  • CEO: "We need an omnichannel experience. Customers should see accurate store inventory online."
  • CFO: "We must shift from CapEx to OpEx. No more buying hardware."
  • CTO: "We want to move to microservices, but we cannot retire the mainframe for at least 3 years due to complex legacy dependencies."

Technical Requirements:

  • Hybrid architecture connecting GCP and on-premises
  • Microservices architecture for the new e-commerce platform
  • PCI-DSS compliance for all payment processing
  • Consistent management plane across on-prem and cloud

Constraints:

  • Mainframe must remain on-premises
  • E-commerce migration must be completed before the next holiday season (8 months)

QUESTION:
To handle the Black Friday traffic (10x normal load), the operations team is concerned about the GKE cluster scaling fast enough. What combination of GCP features should you implement to ensure the platform remains responsive?

Answer options:

A.

Contact Google Cloud Support 24 hours in advance to pre-warm the Global Load Balancer.

B.

Implement Cloud CDN to cache static assets, and configure GKE Cluster Autoscaler alongside Horizontal Pod Autoscaling (HPA).

C.

Over-provision the GKE cluster with 10x the normal node count permanently to ensure capacity.

D.

Use Cloud Run instead of GKE, as GKE cannot scale fast enough for Black Friday traffic.

How to approach this question

Identify the mechanisms for scaling in GCP. CDN reduces backend load. HPA scales pods. Cluster Autoscaler scales nodes. Note that GCP Load Balancers do not need pre-warming.

Full Answer

B.Implement Cloud CDN to cache static assets, and configure GKE Cluster Autoscaler alongside Horizontal Pod Autoscaling (HPA).✓ Correct
Implement Cloud CDN to cache static assets, and configure GKE Cluster Autoscaler alongside Horizontal Pod Autoscaling (HPA).
Handling massive retail spikes requires defense in depth. Cloud CDN absorbs a huge percentage of the traffic (images, CSS, JS) at the edge. For dynamic traffic that reaches the backend, Horizontal Pod Autoscaler (HPA) creates new pods, and the Cluster Autoscaler provisions new underlying VMs to host those pods.

Common mistakes

Choosing option A is a very common mistake for architects coming from AWS, where load balancers historically required pre-warming. GCP's Maglev load balancing architecture handles instant spikes without pre-warming.

Practice the full GCP Professional Cloud Architect Practice Exam 3

50 questions · hints · full answers · grading

More questions from this exam