Examkin

Question

Difficulty: MediumPlanning and Assessing Cloud Compute Solutions

An enterprise financial analytics company is designing a Google Cloud compute strategy for two new application workloads:

1. Workload 1: A stateless web API that experiences unpredictable traffic spikes throughout the day and requires zero infrastructure management overhead along with automatic scale-to-zero capabilities during idle periods.
2. Workload 2: A 3-hour nightly transactional risk batch processing job that is stateless, fully fault-tolerant, and designed to resume smoothly if interrupted.

Which TWO compute resource deployment strategies should the Cloud Engineer recommend to satisfy these technical requirements while optimizing overall compute costs? (Select TWO.)

  1. Deploy Workload 1 on Cloud Run to support automatic scaling down to zero instances during idle periods.Answer
  2. Deploy Workload 2 on Compute Engine using Spot Virtual Machines (VMs) to minimize processing costs.Answer
  3. C
    Deploy Workload 1 on a Compute Engine Managed Instance Group with a fixed node count to eliminate autoscaling configuration latency.
  4. D
    Deploy Workload 2 on Cloud Functions using event triggers to run the continuous 3-hour batch execution.
  5. E
    Deploy Workload 2 on GKE Autopilot with standard node pricing to prevent preemptions, as batch processing cannot utilize Spot instances.

Answer

Deploying Workload 1 on Cloud Run and Workload 2 on Compute Engine using Spot VMs provides the optimal architectural and cost-effective solution.
Cloud Run is the optimal serverless execution engine for stateless web APIs requiring zero management and automatic scaling to zero when idle. For the fault-tolerant 3-hour batch processing job, Compute Engine Spot VMs provide up to 90% cost savings while accommodating potential interruptions.

Step-by-Step Solution

1
Analyze Workload 1 requirements
Identified stateless web API needing zero infrastructure management, auto-scaling up to handle spikes, and scaling to zero when idle.
Cloud Run natively provisions containerized stateless applications, handles HTTP autoscaling, and reduces cost to zero during periods without traffic.
2
Analyze Workload 2 requirements
Identified a 3-hour long-running batch job that is fault-tolerant and stateless.
Because the batch job runs for 3 hours, it exceeds short-lived serverless function limits. Because it is fault-tolerant, Compute Engine Spot VMs offer massive cost discounts (up to 90%) while fulfilling the job requirement.
3
Evaluate and eliminate non-optimal alternatives
Fixed MIGs waste money during idle times, Cloud Functions timeout before 3 hours, and standard GKE Autopilot nodes incur higher baseline costs for a workload that tolerates preemptions.
Matching compute options directly to workload lifecycle and fault tolerance ensures optimal financial and operational trade-offs.

Key Concept

Selecting GCP compute services based on statefulness, execution duration, auto-scaling characteristics, and fault tolerance to optimize cost and management overhead.
Rate this question

Topics

Planning and Assessing Cloud Compute Solutions

More questions from this topic