Examkin

Question

Difficulty: MediumPlanning and Assessing Cloud Compute Solutions

A financial institution is designing compute infrastructure on Google Cloud for two distinct sub-components of a credit risk platform:

1. Workload 1: A stateless, containerized HTTP web service that experiences unpredictable request spikes during trading hours and no traffic overnight. Minimizing costs during idle periods by scaling to zero is a critical requirement.
2. Workload 2: A batch processing job that executes highly parallelized risk simulations. The job is fault-tolerant and can easily resume if an underlying instance is suddenly terminated.

Which TWO compute platform strategies should you recommend to meet these requirements efficiently? (Select TWO)

  1. Deploy Workload 1 on Cloud Run to automatically scale container instances dynamically with HTTP traffic and scale down to zero when idle.Answer
  2. Provision Compute Engine Spot VMs for Workload 2 to run the parallel batch simulation tasks at a significantly lower cost.Answer
  3. C
    Deploy Workload 1 on Cloud Functions by refactoring the containerized application into individual event-driven functions.
  4. D
    Provision standard, non-preemptible Compute Engine VMs with 3-year Committed Use Discounts for Workload 2 to avoid job interruptions.
  5. E
    Deploy Workload 1 on a GKE Autopilot cluster configured with custom OS kernel modules to enable scale-to-zero container execution.

Answer

Deploy Workload 1 on Cloud Run for stateless container scaling to zero, and provision Compute Engine Spot VMs for Workload 2 to execute fault-tolerant batch workloads cost-effectively.
Cloud Run is the optimal compute choice for stateless HTTP container workloads needing automatic scaling down to zero during idle periods. Compute Engine Spot VMs provide maximum cost optimization for batch compute jobs that are resilient to node interruptions.

Step-by-Step Solution

1
Analyze requirements for Workload 1
Identified a stateless, containerized HTTP web service requiring automatic scaling based on HTTP traffic including scaling down to zero.
Cloud Run natively supports container deployments, HTTP autoscaling, and scale-to-zero behavior out of the box.
2
Analyze requirements for Workload 2
Identified a fault-tolerant, parallelized batch processing workload capable of handling node terminations.
Spot VMs provide significant cost reductions for stateless or fault-tolerant batch processing that can tolerate interruptions.

Key Concept

Selecting optimal GCP compute services based on containerization, scaling characteristics, cost sensitivity, and fault tolerance.
Rate this question

Topics

Planning and Assessing Cloud Compute Solutions

More questions from this topic