Examkin

Question

Difficulty: MediumPlanning and Assessing Cloud Compute Solutions

An enterprise logistics company needs to deploy a legacy image processing application on Google Cloud. The application requires specific custom kernel modules and low-level Linux operating system modifications that prevent it from being containerized. The application must run continuously 24 hours a day, 7 days a week, with a consistent workload footprint. Which compute deployment and cost management strategy should you recommend?

  1. Deploy the application on standard Compute Engine virtual machines (VMs) and purchase a 1-year or 3-year Committed Use Discount (CUD) for the compute resources.Answer
  2. B
    Deploy the application on Compute Engine Spot VMs configured within a Managed Instance Group to minimize baseline compute expenses.
  3. C
    Deploy the workload onto Google Kubernetes Engine (GKE) Autopilot clusters to eliminate underlying node management overhead.
  4. D
    Refactor and host the application logic using Cloud Functions with HTTP triggers to enable automatic scaling to zero.

Answer

Deploy the application on standard Compute Engine virtual machines (VMs) and purchase a 1-year or 3-year Committed Use Discount (CUD) for the compute resources.
Compute Engine Virtual Machines provide full OS access, allowing custom Linux kernel configurations that serverless or fully managed container platforms disallow. Because the workload runs continuously 24/7 with a predictable footprint, securing a 1-year or 3-year Committed Use Discount provides significant pricing reduction while guaranteeing instance availability.

Step-by-Step Solution

1
Analyze OS and architectural requirements
The workload requires custom Linux kernel modules and OS-level modifications without containerization capabilities.
Managed container services (GKE Autopilot, Cloud Run) and serverless functions (Cloud Functions) do not allow kernel modifications or uncontainerized host OS access.
2
Select appropriate compute service
Compute Engine Virtual Machines (VMs) must be selected.
Compute Engine provides complete OS-level control, custom image support, and kernel module installation rights.
3
Evaluate workload pattern and pricing strategy
The application runs continuously 24/7 with steady resource usage, favoring Committed Use Discounts over Spot VMs.
Spot VMs risk unexpected terminations for non-fault-tolerant workloads, whereas Committed Use Discounts offer guaranteed availability with substantial cost reductions for steady-state usage.

Key Concept

Selecting GCP Compute Engine options and pricing models based on OS customization constraints and steady-state 24/7 runtime profiles.
Estimated Time:1m 30s
Rate this question

Topics

Planning and Assessing Cloud Compute Solutions

More questions from this topic