mediumMultiple ChoiceObjective-mapped
Google ACE Practice Question: A batch processing job runs on preemptible VMs in…
A batch processing job runs on preemptible VMs in a managed instance group. The job frequently fails due to preemption. Which design change would most effectively improve the job's resilience?
⚠ Common exam trap
Google Cloud often tests the misconception that committed use discounts or sole-tenant nodes provide preemption protection, when in fact they only affect pricing or hardware isolation, not the preemptible VM lifecycle.
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
✓
Use a managed instance group with distribution across multiple zones and enable autoscaling.
Distributing the managed instance group across multiple zones and enabling autoscaling ensures that when preemptible VMs are terminated in one zone, the autoscaler can provision replacement VMs in another zone that still has capacity. This architecture leverages the fact that preemption events are often zone-specific, so multi-zone distribution combined with autoscaling provides resilience without requiring persistent resources.
Answer analysis
Option-by-option breakdown
For each option: why learners choose it and why it is or isn't the right answer here.
- ✗
Use committed use discounts (1-year or 3-year).
Why it's wrong here
Committed use discounts (CUDs) offer a pricing discount in exchange for a 1- or 3-year spend commitment, but they do nothing to change the fundamental behavior of preemptible VMs: Compute Engine can still reclaim them at any time. CUDs apply to on-demand or committed-use VM configurations, not to preemptible VMs, and they don't add redundancy, failover, or instance replacement logic. A batch job would still be interrupted just as frequently, and you'd be locked into paying for capacity that may be repeatedly preempted, so this option fails to address the resilience requirement.
- ✗
Add GPUs to the instances.
Why it's wrong here
Adding GPUs to preemptible VMs is counterproductive because GPUs themselves are a scarce resource and preemptible VMs with GPUs have even lower availability and higher preemption rates than standard preemptible instances. Moreover, not all GPU types are supported on preemptible VMs, and attaching a GPU increases the chance that the instance cannot be scheduled at all, making the batch job less reliable. This option adds cost and complexity without providing any mechanism to detect failed instances, replace them, or spread work across zones, so it does not solve the preemption problem.
- ✗
Use sole-tenant nodes.
Why it's wrong here
Sole-tenant nodes dedicate a physical server to your project, which gives you hardware isolation and predictable capacity, but they do not protect against preemption—preemptible VMs on sole-tenant nodes are still subject to the same 24-hour maximum lifetime and can be terminated by Compute Engine at any time. In fact, sole-tenant nodes are designed for workloads that require compliance or specialized hardware, not for improving fault tolerance to spot-style reclaiming. They also don't include any automatic healing or cross-zone failover, so a batch job running on sole-tenant nodes would still stall when a preemptible VM is terminated.
- ✓
Use a managed instance group with distribution across multiple zones and enable autoscaling.
Why this is correct
A managed instance group (MIG) with instances distributed across multiple zones directly mitigates preemption by ensuring that a single zone's preemption event doesn't wipe out the entire batch capacity; MIGs also provide automatic instance replacement via health checks and instance re-creation. Enabling autoscaling lets the group scale the number of preemptible VMs based on job demand or queue depth, and you can even configure the MIG to use preemptible VMs as the default while maintaining a buffer of on-demand instances if needed. This combination creates a resilient, self-healing architecture that tolerates preemption events and keeps the batch job making progress, which is exactly what the question requires.
Go deeper
Related to this question
Learn chapter
Persistent Disk Types: SSD, Balanced, Standard
Key term
Managed instance group
A managed instance group is a collection of identical virtual machine instances that are automatically managed as a single unit to ensure high availability and scalability.
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.
About these practice questions
One of 769 original ACE practice questions on Courseiva, each with a full explanation and wrong-answer analysis — not exam dumps or protected exam content. Learn why practice questions differ from exam dumps →
JA
Written by Johnson Ajibi, MSc IT Security
Senior Network & Security Engineer · founder of Courseiva
This ACE 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 ACE exam.