Examkin

Question

Difficulty: HardPlanning Compute Engine Resources and Machine Types

A security infrastructure team is designing a Google Compute Engine architecture for a 24/7 continuous log parsing engine. Profiling shows that the workload requires exactly 6 vCPUs6\text{ vCPUs} and 45 GB45\text{ GB} of RAM during steady-state operations. Standard machine types either under-allocate RAM or over-allocate vCPUs, leading to unnecessary operational expenses. The application cannot tolerate preemption or unexpected instance termination. Which configuration and purchasing strategy should the cloud engineer recommend to minimize cost while meeting all technical requirements?

  1. Provision a Compute Engine instance using a custom machine type with 6 vCPUs and 45 GB of RAM, and purchase Committed Use Discounts (CUDs) for the required vCPU and memory capacity.Answer
  2. B
    Provision Spot VM instances configured with a custom machine type of 6 vCPUs and 45 GB of RAM to achieve the highest per-hour cost savings without contractual commitments.
  3. C
    Provision standard N2 machine instances with 8 vCPUs and 64 GB of RAM, relying entirely on automatic Sustained Use Discounts (SUDs) to achieve maximum cost savings.
  4. D
    Migrate the continuous long-running log parsing engine to event-driven Cloud Functions to eliminate instance sizing decisions entirely.

Answer

Provision a Compute Engine instance using a custom machine type with 6 vCPUs and 45 GB of RAM, combined with Committed Use Discounts for the baseline vCPU and RAM capacity.
The combination of custom machine types and Committed Use Discounts (CUDs) directly addresses both requirements: custom sizing eliminates waste from unused vCPUs by configuring exactly 6 vCPUs and 45 GB RAM, while CUDs offer the highest price reduction for steady 24/7 non-interruptible workloads.

Step-by-Step Solution

1
Analyze resource sizing requirements.
The workload needs a specific ratio of 6 vCPUs6\text{ vCPUs} to 45 GB45\text{ GB} of RAM (7.5 GB per vCPU7.5\text{ GB per vCPU}), which does not fit standard predefined machine types without over-provisioning.
Custom machine types in Compute Engine allow tailoring vCPU and RAM counts precisely, preventing wasteful spending on unused vCPUs.
2
Evaluate workload availability requirements.
The application runs continuous 24/7 workloads and cannot tolerate unexpected termination.
Spot VMs are fault-tolerant/batch target instances that can be reclaimed at any time, making them unsuitable for continuous uninterrupted processing.
3
Determine the optimal discount strategy.
For predictable 24/7 baseline capacity, Committed Use Discounts (1-year or 3-year) offer significantly higher savings than automatic Sustained Use Discounts.
CUDs provide up to 57% to 70% savings on committed vCPU and memory resources for continuous workloads.

Key Concept

Compute Engine Custom Machine Types and Committed Use Discount Planning
Estimated Time:2m 0s
Rate this question

Topics

Planning Compute Engine Resources and Machine Types

More questions from this topic