NCP-AIO Workload Management Practice Question
An AI operations engineer manages a shared Kubernetes cluster where several teams submit GPU jobs. The engineer must prevent any single namespace from consuming all GPU capacity and must also ensure that jobs from one team cannot starve others during peak periods. Which combination of Kubernetes and NVIDIA GPU Operator features should the engineer implement?
⚠ Common exam trap
The trap here is treating NetworkPolicy, taints, or LimitRange as tools for GPU capacity fairness, when only ResourceQuota enforces an aggregate per-namespace ceiling on nvidia.com/gpu.
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
✓
Define a ResourceQuota on nvidia.com/gpu per namespace and use the GPU Operator's device plugin to advertise capacity so the quota can enforce limits.
ResourceQuota is the Kubernetes mechanism that caps aggregate resource consumption per namespace, and it works for GPU extended resources once the NVIDIA device plugin advertises them. By setting a quota on nvidia.com/gpu, the administrator prevents any single namespace from consuming all GPUs, ensuring other teams retain capacity. This directly addresses both the monopolization and starvation concerns.
Answer analysis
Option-by-option breakdown
For each option: why learners choose it and why it is or isn't the right answer here.
- ✓
Define a ResourceQuota on nvidia.com/gpu per namespace and use the GPU Operator's device plugin to advertise capacity so the quota can enforce limits.
Why this is correct
ResourceQuota can cap the total nvidia.com/gpu requests within a namespace when the device plugin advertises GPUs as extended resources. This prevents one namespace from monopolizing cluster GPU capacity. Combined with the operator's device plugin, quotas become enforceable at admission time, ensuring fair sharing across teams without manual intervention.
- ✗
Enable the MIG Manager and assign a fixed mig profile to every namespace through a LimitRange.
Why it's wrong here
LimitRange sets default and maximum resource values per pod or container, but it does not aggregate usage across a namespace the way ResourceQuota does. Assigning MIG profiles also changes resource names and partitioning, which is a different isolation model. This combination does not enforce a namespace-wide GPU ceiling, so it fails the stated requirement.
- ✗
Apply a NetworkPolicy that restricts pod-to-pod traffic between namespaces and rely on it to limit GPU consumption.
Why it's wrong here
NetworkPolicy governs network traffic, not compute resource consumption. It cannot cap how many GPUs a namespace requests or prevent a team from exhausting cluster GPU capacity. While useful for security isolation, it has no mechanism to enforce GPU quotas, so it fails to address the starvation requirement in this scenario.
- ✗
Set a node taint for nvidia.com/gpu and add tolerations only to the highest-priority team's pods.
Why it's wrong here
Taints and tolerations control which pods may schedule onto which nodes, but they do not impose per-namespace capacity ceilings. If multiple teams have tolerations, none is prevented from consuming all GPUs. This approach also risks leaving GPUs idle when only one team can schedule, so it does not achieve fair, quota-based sharing.
About these practice questions
This NCP-AIO question is part of Courseiva's 309-question bank — original exam-style content with full explanations and wrong-answer analysis, never real exam questions or exam 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 NVIDIA exam blueprint
This NCP-AIO practice question is part of Courseiva's free NVIDIA 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 NCP-AIO exam.