Examkin

Soru

Zorluk: ZorCapacity Planning and Infrastructure Workload Scaling Optimization

A financial analytics enterprise processes real-time risk assessment calculations on Google Cloud using a Compute Engine Managed Instance Group (MIG) deployed across multiple zones in `us-central1`. During unpredictable market volatility events, compute demand spikes rapidly, requiring the MIG to scale up from 20 to 150 `n2-standard-16` instances. During a recent market event, autoscaling failed to provision required instances due to reaching the region's N2 vCPU quota limit and encountering transient stockouts of `n2-standard-16` instances in specific zones. The enterprise requires a solution that guarantees instance allocation during market spikes while adhering to GCP quota governance. Which architectural strategy should the Cloud Architect recommend?

  1. Submit a proactive regional N2 vCPU quota increase request for `us-central1` and configure Compute Engine zonal capacity reservations for the required `n2-standard-16` capacity.Cevap
  2. B
    Migrate the risk assessment engine to a Google Kubernetes Engine (GKE) cluster using Cluster Autoscaler, expecting GKE to automatically bypass regional vCPU quota limits during rapid scaling.
  3. C
    Replace the Compute Engine instances with a stateless Cloud Run service backed by Cloud Spanner to absorb compute bursts without needing quota management.
  4. D
    Establish an HA VPN connection to burst traffic to an on-premises data center during peak market events to circumvent cloud capacity limits.

Cevap

The Cloud Architect should submit a proactive regional N2 vCPU quota increase for `us-central1` and configure zonal capacity reservations for `n2-standard-16` instances.
To reliably handle scale-up events from 20 to 150 instances, both administrative quota and physical resource availability must be secured. Submitting a proactive regional vCPU quota increase for N2 instances in `us-central1` resolves project quota restrictions. Creating Compute Engine capacity reservations ensures Google Cloud physically reserves the required `n2-standard-16` compute capacity in the target zones, enabling the Managed Instance Group to scale seamlessly without running into stockouts.

Adım Adım Çözüm

1
Analyze capacity failure root causes
Identified two bottlenecks: regional vCPU project quota limit and hardware availability/stockouts in specific compute zones.
Scaling limits in Compute Engine occur at both administrative quota levels and physical hardware resource availability levels.
2
Evaluate GCP Quota Management
Proactively requesting a regional vCPU quota increase ensures project quota does not block instance creation during scale-up.
Quota limits exist to prevent unexpected spend and ensure fair resource distribution, requiring explicit quota requests prior to high-demand events.
3
Evaluate Infrastructure Reservation
Configuring Compute Engine Zonal Capacity Reservations guarantees that specific machine types (`n2-standard-16`) are reserved in the designated zones.
Capacity reservations ensure resources are physically set aside in target zones so autoscaling MIGs can consume them reliably during burst events.

Anahtar Kavram

Capacity Planning and Infrastructure Workload Scaling Optimization
Bu soruyu puanla

Konular

Capacity Planning and Infrastructure Workload Scaling Optimization