Google PCA Design and plan a cloud solution architecture Practice Question
A retail company runs a batch analytics workload on Compute Engine. Jobs run nightly, are fault-tolerant, and can be preempted. Finance wants to minimize compute cost while ensuring the jobs still complete each night. The jobs are managed by a Managed Instance Group (MIG) template that must stay within a single zone for data locality compliance. Which configuration should you recommend?
⚠ Common exam trap
The trap here is assuming that committed use discounts are always the cheapest option, when they only pay off for steady, long-running workloads rather than nightly preemptible batch jobs.
Answer choices
Why each option matters
Answer the question above first, then reveal the full breakdown to understand why each option is right or wrong.
Correct answer & explanation
✓
Create a zonal MIG with a spot VM instance template and set the autoscaler to scale based on CPU utilization.
Spot VMs offer the largest discount for fault-tolerant, preemptible workloads, and a zonal MIG preserves the single-zone data locality requirement. The autoscaler ensures the nightly job has enough instances to finish on time. Committed use and sustained use discounts target steady-state or long-running workloads, so they do not fit a short nightly batch window. The combination of spot VMs, zonal placement, and CPU-based autoscaling meets both the cost and compliance constraints.
Answer analysis
Option-by-option breakdown
For each option: why learners choose it and why it is or isn't the right answer here.
- ✗
Create a zonal MIG with E2 standard VMs and configure a sustained use discount to lower the price.
Why it's wrong here
Sustained use discounts apply automatically to qualifying on-demand usage but only after a VM runs for a significant portion of the month, which nightly jobs will not reach. E2 standard VMs cost more than spot VMs for the same capacity, so this fails to minimize cost. The zoning is correct, but the pricing mechanism and machine choice are not optimal for preemptible batch work.
- ✗
Create a regional MIG with committed use discounts applied to the template and scale across three zones.
Why it's wrong here
A regional MIG spreads instances across multiple zones, which violates the single-zone data locality compliance requirement. Committed use discounts reduce cost only when you commit to steady-state usage, which does not fit a nightly batch pattern, and they do not help with preemptible capacity. This option is wrong because it breaks the zoning constraint and uses an inappropriate discount model.
- ✗
Create a zonal MIG with custom machine types and use the committed use discount for one-year terms.
Why it's wrong here
Committed use discounts require a one- or three-year commitment for a specific amount of vCPU and memory, which is wasteful for a workload that only runs at night. Custom machine types can reduce cost slightly but do not approach the discount of spot VMs. This option fails because the commitment model does not match the intermittent batch schedule and the savings are far smaller.
- ✓
Create a zonal MIG with a spot VM instance template and set the autoscaler to scale based on CPU utilization.
Why this is correct
Spot VMs provide up to 60-91% discount versus standard VMs and are ideal for fault-tolerant, preemptible batch jobs. A zonal MIG keeps instances in one zone, satisfying the data locality requirement, and the autoscaler adds capacity when CPU rises during the nightly run. The job must tolerate preemption, which it does, so spot VMs directly minimize cost without violating the stated constraints.
Go deeper
Related to this question
Learn chapter
Load Balancing and Autoscaling
Key term
Autoscaler
An Autoscaler is a cloud service that automatically increases or decreases the number of virtual machines (instances) or resources based on real-time demand, so your application always has enough capacity without wasting money on idle servers.
Key term
Batch
Batch is a cloud computing service that runs large numbers of computing jobs as a group, or batch, without needing to manage individual servers.
About these practice questions
Courseiva writes every PCA question from scratch — 807 in total, each with an explanation and a wrong-answer breakdown. None are copied from real exams or dumps. Learn why practice questions differ from exam dumps →
JA
Written and reviewed by Johnson Ajibi, MSc IT Security
Senior Network & Security Engineer · founder of Courseiva
Last reviewed September 2026 · checked against the official Google Cloud exam blueprint
This PCA practice question is part of Courseiva's free Google Cloud certification practice question bank. Courseiva provides original exam-style practice questions with explanations, topic-based practice, mock exams, readiness tracking, and study analytics to help learners prepare for the PCA exam.