A logistics enterprise executes a weekly batch reporting job every Saturday that takes 3 hours to complete. The application is designed to be fault-tolerant and can resume processing from saved checkpoints if instance preemption occurs. The business objective is to minimize compute infrastructure costs for this short-lived, fault-tolerant workload without long-term financial commitments. Which Google Cloud compute strategy should the cloud architect recommend?
- Provision Compute Engine Spot VMs to run the batch reporting jobs.Answer
- BPurchase 3-year Committed Use Discounts (CUDs) to cover the batch compute capacity.
- CDeploy a multi-zone Google Kubernetes Engine (GKE) Standard cluster running continuous 24/7 node pools.
- DProvision a multi-region Cloud Spanner instance to compute and store the batch metrics continuous data.
Answer
Provisioning Compute Engine Spot VMs is the most cost-effective solution for short-duration, fault-tolerant batch workloads.
Spot VMs leverage spare Google Cloud capacity at up to a 60–91% discount compared to standard on-demand instances. Because the batch job is fault-tolerant and runs for only 3 hours per week, Spot VMs fulfill the business requirement of minimizing cost without introducing long-term billing commitments.
Step-by-Step Solution
Key Concept
Compute Engine Spot VMs for cost-optimized fault-tolerant batch processing