Examkin

Question

Difficulty: MediumPlanning Compute Engine Resources and Machine Types

An enterprise is planning to migrate a critical, stateful relational database workload to Google Cloud Compute Engine. The database requires persistent storage, continuous uptime, and predictable performance without abrupt shutdowns. Which Compute Engine resource planning strategy should you select?

  1. Deploy the database on standard Compute Engine N2 instances using persistent disks and apply Committed Use Discounts for cost optimization.Answer
  2. B
    Deploy the database on Spot VMs using custom machine types to minimize compute costs.
  3. C
    Containerize the relational database and host it on Cloud Run to achieve zero server management overhead.
  4. D
    Deploy the database workload on GKE Autopilot using ephemeral local storage for database files.

Answer

Deploy the database on standard Compute Engine N2 instances using persistent disks and apply Committed Use Discounts for cost optimization.
Standard Compute Engine instances provide guaranteed resource availability and continuous execution necessary for stateful database workloads. Combining standard instances with persistent disks and Committed Use Discounts (CUDs) fulfills SLA requirements while optimizing costs for predictable, long-running compute workloads.

Step-by-Step Solution

1
Analyze workload requirements
The workload is a stateful database requiring uninterrupted operations and persistent storage.
Databases cannot tolerate unexpected terminations without risking data corruption or downtime.
2
Evaluate VM machine options and lifecycle settings
Standard instances (non-preemptible/non-Spot) guarantee uptime availability and support persistent disks.
Spot VMs are subject to abrupt preemption and should only be used for fault-tolerant batch workloads.
3
Select cost optimization strategy suitable for steady-state workloads
Committed Use Discounts offer significant savings for predictable continuous usage.
Committed Use Discounts provide price reductions for steady-state workloads without risking instance availability.

Key Concept

Planning Compute Engine Instance Types and Lifecycle for Stateful Workloads
Rate this question

Topics

Planning Compute Engine Resources and Machine Types

More questions from this topic