Examkin

Question

Difficulty: HardOptimizing Business Processes through FinOps and Cloud Cost Management

A high-frequency fintech platform operates its core transaction processing engine on Google Cloud using Compute Engine managed instance groups alongside stateless validation microservices running on Cloud Run. The core transaction processing engine maintains a constant, predictable 24/7 baseline usage, whereas the validation microservices undergo unpredictable 10x traffic bursts during sudden financial market fluctuations. The platform engineering team also executes ad-hoc analytical queries against BigQuery to monitor fraud patterns. Which cost optimization strategy should a Cloud Architect recommend to maximize spend efficiency without sacrificing operational availability?

  1. Purchase spend-based Flexible Committed Use Discounts (CUDs) to cover the predictable baseline compute usage across Compute Engine and Cloud Run, while utilizing auto-scaling with Spot VMs for stateless burst capacity and standard on-demand slots for BigQuery ad-hoc queries.Answer
  2. B
    Purchase 3-year resource-based Committed Use Discounts (CUDs) sized to cover peak anticipated vCPU and memory requirements across both Compute Engine and Cloud Run to lock in maximum hourly discount rates for all traffic spikes.
  3. C
    Migrate all stateless Cloud Run validation microservices to a dedicated Google Kubernetes Engine (GKE) cluster provisioned with 3-year committed use node reservations to consolidate container workloads and eliminate serverless request pricing overhead.
  4. D
    Grant primitive Billing Account Administrator and Owner IAM roles to engineering team leads in order to let individual project teams directly manage and purchase resource commitments independently.

Answer

Purchase spend-based Flexible Committed Use Discounts (CUDs) to cover the predictable baseline compute footprint across Compute Engine and Cloud Run, while utilizing auto-scaling with Spot VMs for stateless burst capacity and standard on-demand slots for BigQuery ad-hoc queries.
The solution correctly identifies that spend-based Flexible Committed Use Discounts (CUDs) should be applied to predictable, steady-state baseline usage across Compute Engine and Cloud Run. For erratic, high-volume traffic spikes, leveraging auto-scaling with Spot VMs or pay-as-you-go capacity ensures high cost efficiency without over-committing capital. Furthermore, keeping BigQuery on on-demand slot allocation avoids paying for idle reserved slots when queries are infrequent.

Step-by-Step Solution

1
Analyze workload patterns to separate predictable baseline consumption from bursty, variable consumption.
Identified constant 24/7 baseline compute across Compute Engine and Cloud Run, alongside erratic 10x traffic spikes.
FinOps best practices require aligning commitment models only to stable, non-variable baseline requirements.
2
Select the appropriate Google Cloud discount commitment mechanism for multi-service compute baselines.
Chosen spend-based Flexible CUDs, which offer hourly spend flexibility across Compute Engine and Cloud Run.
Flexible CUDs apply automatically across multiple compute products without tying the discount to specific machine types or regions.
3
Select cost-efficient mechanisms for stateless, unpredictable burst capacity and ad-hoc analytics.
Applied Spot VMs for fault-tolerant burst capacity and on-demand pricing for ad-hoc BigQuery queries.
Spot VMs provide significant cost reductions for batch/burst microservices without long-term monetary commitments.

Key Concept

FinOps Cost Governance and Flexible Committed Use Discount (CUD) Optimization
Estimated Time:2m 0s
Rate this question

Topics

Optimizing Business Processes through FinOps and Cloud Cost Management