Examkin

Soru

Zorluk: ZorCapacity Planning and Infrastructure Workload Scaling Optimization

An enterprise financial services provider runs a payroll processing platform on Google Cloud using Compute Engine Managed Instance Groups (MIGs). The application experiences predictable, severe 10x traffic bursts on the last business day of every month. During the most recent burst, instances failed to scale out despite autoscaling policies triggering, resulting in request timeouts due to hitting regional Compute Engine vCPU quotas. The organization requires a solution that guarantees instance availability for scheduled peak events without incurring continuous baseline compute costs. Which capacity planning and scaling strategy should the principal cloud architect recommend?

  1. Submit a regional vCPU quota increase request in advance and implement targeted Compute Engine On-Demand Reservations scheduled for the monthly peak window.Cevap
  2. B
    Migrate the workloads from Compute Engine MIGs to Google Kubernetes Engine (GKE) with Cluster Autoscaler to automatically bypass regional vCPU quota limits.
  3. C
    Reconfigure the Compute Engine MIG autoscaling policy to lower the target CPU utilization threshold from 80% to 30% so scaling actions trigger earlier during traffic bursts.
  4. D
    Migrate the application's underlying database tier to Cloud Spanner across multiple regions to resolve compute node scaling constraints.

Cevap

The correct strategy is to request a regional vCPU quota increase in advance and leverage scheduled Compute Engine On-Demand Reservations during the monthly processing window.
Requesting a regional vCPU quota increase in advance ensures the GCP project has sufficient quota headroom for peak demand. Combining this with Compute Engine Reservations guarantees that the physical infrastructure capacity is reserved in the target zone for the predictable burst window, preventing capacity exhaustion while avoiding continuous baseline compute expenditure.

Adım Adım Çözüm

1
Analyze the workload failure root cause
Identified that scaling failed due to hitting regional vCPU quota limits and lacking guaranteed compute capacity during predictable traffic bursts.
Autoscaling policies cannot provision resources beyond defined GCP quota ceilings or when zonal capacity is exhausted.
2
Address project quota boundaries
Submit a quota increase request in the GCP Console prior to peak events.
Quota increases require lead time for approval and execution, preventing scaling failures caused by hard limit breaches.
3
Ensure capacity availability cost-effectively
Configure Compute Engine Reservations for the required capacity during the processing window.
Reservations guarantee that hardware resources are available in the target zone during peak windows without requiring instances to remain running continuously outside the peak window.

Anahtar Kavram

Capacity Planning and Infrastructure Workload Scaling Optimization
Bu soruyu puanla

Konular

Capacity Planning and Infrastructure Workload Scaling Optimization