Prepare and test your skills

Prepare and test your skills

Free sample question with answers and detailed explanation below. The Cloud Digital Leader practice bank has an additional 3,070 practice questions.Try 10 free realistic questions: instant score, no credit card, and no account required.
An e-commerce company runs a stateless web application on Compute Engine virtual machines (VMs). Recently, unpredictable traffic surges caused high CPU utilization and slow response times, while quiet periods left extra VMs sitting idle. Additionally, during a recent traffic spike, the application software crashed on several VMs even though the VMs remained running, resulting in failed customer requests.
The company wants to update its architecture to meet the following requirements:
Which Google Cloud solution should the company implement?
Committed Use Discounts (CUDs) are a billing cost-optimization mechanism that provides discounted prices in exchange for committing to a consistent level of resource usage (such as vCPUs and memory) for a one-year or three-year term. Rate quotas are per-project service limits that restrict the number of API requests a system can receive within a specific period to control resource consumption and protect against excessive load.
This approach would be appropriate if the company ran a predictable, steady-state workload where it wanted to lower long-term compute bills using Committed Use Discounts, and needed to protect backend APIs from abusive request floods or runaway client scripts using rate quotas.
A common misconception is equating cost optimization through billing discounts with resource optimization through elasticity. While CUDs reduce the unit price of compute resources, autoscaling optimizes actual resource consumption by dynamically removing idle instances when demand is low.
Spot VMs are Compute Engine virtual machines that provide access to unused Google Cloud compute capacity at a steep discount compared to standard VMs, with the trade-off that Google Cloud can preempt (terminate) them at any time if the capacity is needed elsewhere. Cloud Billing budgets and alerts are Cost Management tools that track cloud spending against defined financial thresholds and send notifications to administrators when costs approach or exceed the budget.
This strategy would be appropriate if the company wanted to run fault-tolerant, non-urgent batch processing jobs at the lowest possible compute price using Spot VMs, while keeping finance teams informed of monthly project expenditures using Cloud Billing budgets and alerts.
Students often confuse financial alerting with operational autoscaling. While budgets notify teams about cost increases, maintaining performance continuity under variable load requires elastic infrastructure—such as a Managed Instance Group (MIG) with an autoscaling policy—that acts immediately without human intervention.
Compute Engine capacity reservations provide assured hardware capacity in a specific zone for a designated number of VMs of a chosen machine type, ensuring resources are available when needed. Recommender (part of Active Assist) is an analytical tool that evaluates historical resource utilization data to generate intelligent recommendations for right-sizing underutilized or over-provisioned cloud resources.
This solution would be ideal if the company needed guaranteed zonal hardware availability ahead of a planned product launch using capacity reservations, and wanted to review historical utilization trends to right-size baseline VM machine types using Recommender.
While Recommender and Cloud Monitoring are essential for ongoing capacity planning and baseline right-sizing, they complement—rather than replace—automated runtime mechanisms like Managed Instance Groups (MIGs), autoscaling, and autohealing.
A Compute Engine Managed Instance Group (MIG) is a collection of identical virtual machine (VM) instances that you control and scale as a single entity using an instance template. When combined with autoscaling policies, Cloud Load Balancing health checks, and autohealing, a MIG provides an elastic, resilient, and self-healing infrastructure architecture that automatically adapts to changing workload demands and recovers from software disruptions.
RUNNING state, the load balancer stops sending user requests to any backend instance where the application has crashed.
Unlike static provisioning or manual resizing strategies, combining MIG autoscaling, load-balancing health checks, and autohealing addresses both capacity fluctuations and application-level reliability simultaneously. It ensures continuous performance during unexpected surges while keeping operational costs aligned with actual consumption.
Answer them yourself and get your score the moment you finish. No account needed.
Full access includes
Keep the momentum going with these hand-picked practice scenarios
Last updated