An enterprise e-commerce company manages dozens of Google Cloud projects across multiple product engineering teams. The infrastructure consists of steady-state backend microservices running on Compute Engine alongside highly variable, fault-tolerant batch analytical jobs and ad-hoc data exploration. The centralized FinOps team needs to establish organizational cost governance, improve spending visibility, and reduce overall infrastructure expenditure without restricting developer velocity. Which of the following actions should the Cloud Architect recommend? (Select TWO.)
- Enable organization-level Cloud Billing data export to BigQuery and configure Cloud Billing Budgets with Pub/Sub notifications for automated spending alerts and governance.Answer
- Apply Committed Use Discounts (CUDs) to cover predictable, steady-state Compute Engine baseline workloads at the billing account level, and utilize Spot VMs for fault-tolerant batch workloads.Answer
- CPurchase 3-year standard Committed Use Discounts (CUDs) to cover peak resource usage during unpredictable, bursty developer ad-hoc data analysis workloads.
- DMigrate simple stateless web microservices from Cloud Run to dedicated Google Kubernetes Engine (GKE) clusters to eliminate per-request serverless charges.
- EAssign the primitive Owner IAM role on the Billing Account to individual engineering team leads so they can directly purchase commitments and manage billing hierarchies.
Answer
The Cloud Architect should recommend enabling organization-level Cloud Billing data export to BigQuery combined with Cloud Billing budget alerts via Pub/Sub, as well as applying Committed Use Discounts for steady-state baseline workloads while running fault-tolerant batch workloads on Spot VMs.
The correct strategy combines technical right-sizing and financial governance. Exporting organizational billing data to BigQuery provides centralized visibility across all projects and supports custom analysis, while integrating billing budgets with Cloud Pub/Sub allows automated cost control actions. Additionally, applying CUDs to predictable baseline capacity secures significant discounts, while using Spot VMs for fault-tolerant batch jobs reduces compute costs by up to 60-90% without committing to unneeded capacity.
Step-by-Step Solution
Key Concept
FinOps Cloud Cost Optimization and Governance Framework