Examkin

Soru

Zorluk: OrtaPlanning Compute Engine Resources and Machine Types

An e-commerce enterprise is planning its compute architecture on Google Compute Engine for two distinct operational workloads:

1. Core Order API: A mission-critical 24/7 web application requiring continuous uptime and predictable CPU/memory resources without unexpected terminations.
2. Nightly Catalog Indexer: A fault-tolerant batch processing job that runs for 2 hours each night, writes incremental checkpoints to Cloud Storage, and can withstand sudden instance preemption.

Which TWO cost-effective provisioning strategies should the cloud engineer implement? (Select TWO)

  1. Purchase 1-year or 3-year Committed Use Discounts (CUDs) for the standard Virtual Machines running the Core Order API.Cevap
  2. Use Spot Virtual Machines for the Nightly Catalog Indexer batch processing workload.Cevap
  3. C
    Provision Spot Virtual Machines for the Core Order API service to minimize baseline infrastructure costs.
  4. D
    Rely exclusively on automatic Sustained Use Discounts (SUDs) for the Nightly Catalog Indexer batch job rather than using Spot VMs.

Cevap

The cloud engineer should purchase Committed Use Discounts for the continuous Core Order API service and utilize Spot VMs for the fault-tolerant Nightly Catalog Indexer batch workload.
For continuous 24/7 mission-critical services that require guaranteed availability, standard VM instances backed by Committed Use Discounts (CUDs) supply significant cost reduction without risk of interruption. Conversely, batch workloads that run briefly each night and checkpoint their state are fault-tolerant and uniquely suited for Spot VMs, which deliver maximum discounts for preemptible capacity.

Adım Adım Çözüm

1
Evaluate the uptime and interrupt-tolerance requirement for the Core Order API service.
Because the Core Order API is a continuous 24/7 mission-critical service that cannot tolerate abrupt terminations, standard compute instances backed by 1-year or 3-year Committed Use Discounts (CUDs) provide guaranteed availability with optimized pricing.
CUDs offer high discount rates for predictable long-term usage without sacrificing instance reliability.
2
Evaluate the workload characteristics of the Nightly Catalog Indexer job.
Because the indexing job runs for only 2 hours nightly and writes state checkpoints to Cloud Storage, it is fully fault-tolerant and ideal for Spot VMs.
Spot VMs offer steep cost savings for short-lived, interruptible batch processing workloads.

Anahtar Kavram

Planning Compute Engine Machine Types and Pricing Models (CUDs vs Spot VMs vs SUDs)
Bu soruyu puanla

Konular

Planning Compute Engine Resources and Machine Types

Bu konudan daha fazla soru