Soru

Zorluk: KolayCapacity Planning and Infrastructure Workload Scaling Optimization

An e-commerce company is preparing for an upcoming flash sale expected to increase traffic to five times normal peak volume on Google Cloud. As the Cloud Architect responsible for capacity planning and workload scaling optimization, which TWO actions should you recommend to ensure seamless infrastructure scaling and prevent availability bottlenecks? (Select TWO answers.)

  1. Review regional GCP quota limits for CPU cores and IP addresses in advance, submitting quota increase requests prior to the event.Cevap
  2. Configure Compute Engine Managed Instance Group (MIG) autoscaling policies based on custom application metrics such as request latency or HTTP queue depth.Cevap
  3. C
    Migrate the existing single-region Cloud SQL relational database to a multi-region Cloud Spanner instance specifically to handle the temporary event load.
  4. D
    Defer capacity planning configuration changes and rely on GCP default quota limits, assuming compute quotas scale automatically as demand increases.

Cevap

Review regional GCP quota limits for CPU cores and IP addresses in advance, submitting quota increase requests prior to the event, AND configure Compute Engine Managed Instance Group (MIG) autoscaling policies based on custom application metrics such as request latency or HTTP queue depth.
Effective capacity planning requires requesting regional resource quota increases well ahead of anticipated high-volume events because quota requests are not instant. Additionally, configuring scaling policies to utilize custom workload metrics (such as queue depth or HTTP throughput) ensures instance groups scale out responsively when I/O or connection limits are reached.

Adım Adım Çözüm

1
Evaluate project quota boundaries
Identify potential resource constraints (e.g., regional CPU limits) before traffic spikes occur.
Quota increases require administrative approval and time to process.
2
Optimize autoscaling signals
Ensure instance groups scale based on actual load bottlenecks (like incoming HTTP requests or queue length) rather than CPU alone.
Web workloads can saturate request queues or network sockets while CPU utilization remains low.

Anahtar Kavram

Capacity Planning and Infrastructure Workload Scaling Optimization
Tahmini Süre:1m 0s
Bu soruyu puanla