Examkin

Soru

Zorluk: OrtaPlanning Compute Engine Resources and Machine Types

A digital publishing platform is planning its Google Compute Engine infrastructure for two distinct backend services:

1. A high-throughput, in-memory caching tier that requires high memory-to-vCPU ratios, operating 24/7 with a 99.99% uptime requirement.
2. A batch image-resizing pipeline that processes asynchronous worker queues, where individual instance preemptions can be retried automatically without data loss.

Which machine configuration and pricing strategy should the cloud engineer select to optimize compute costs while satisfying the SLA requirements?

  1. Provision high-memory machine family instances backed by Committed Use Discounts for the caching tier, and provision Spot VMs for the batch image-resizing pipeline.Cevap
  2. B
    Provision Spot VMs for the in-memory caching tier to maximize hourly cost savings, and purchase 1-year Committed Use Discounts on standard machine types for the batch image-resizing pipeline.
  3. C
    Migrate the stateful in-memory caching tier to GKE Autopilot clusters to eliminate node-level management, assuming node memory state is preserved when Autopilot scales nodes to zero.
  4. D
    Use standard machine types for both workloads and rely exclusively on automatic Sustained Use Discounts without leveraging Spot VMs or Committed Use Discounts.

Cevap

Provision high-memory machine family instances backed by Committed Use Discounts for the caching tier, and provision Spot VMs for the batch image-resizing pipeline.
The correct strategy combines high-memory machine types with Committed Use Discounts for the 24/7 stateful caching service to guarantee availability and lower baseline costs, while leveraging Spot VMs for the fault-tolerant batch processing pipeline to maximize cost savings on interruptible tasks.

Adım Adım Çözüm

1
Analyze workload characteristics for the caching tier.
Identified as a high-memory, 24/7 continuous production service requiring high availability.
Continuous workloads with high memory requirements benefit from high-memory machine families and Committed Use Discounts (1-year or 3-year) for maximum cost efficiency without sacrificing availability.
2
Analyze workload characteristics for the batch processing pipeline.
Identified as a stateless, fault-tolerant, queue-based batch processing service.
Fault-tolerant batch jobs that handle interruptions seamlessly are ideal candidates for Spot VMs, which offer up to 60-91% discounts compared to standard pricing.
3
Combine the machine family selection with the optimal pricing models.
High-memory instances with CUDs for cache tier + Spot VMs for batch rendering.
This combination minimizes total cost of ownership while fully meeting performance and SLA constraints.

Anahtar Kavram

Compute Engine Machine Type Selection & Discount Models (CUD vs. Spot VMs)
Bu soruyu puanla

Konular

Planning Compute Engine Resources and Machine Types

Bu konudan daha fazla soru