Key details for this exam, checked against the published exam outline
Each question shows the correct answer and an explanation of why it is right
Refer to the exhibit.
A current Nutanix cluster is hosting an NKP cluster and a production Kubernetes cluster named production. Infrastructure administrators recently added three nodes with NVidia GPUs for a new AI initiative, and now a Platform Engineer has been asked to add three workers with the GPUs to the production Kubernetes cluster. Which first step should the engineer take to achieve this task?
A Platform Engineer has been tasked with setting up a secure and isolated environment for managing an NKP environment. The environment will have no access to the Internet but will be required to create additional NKP clusters in the future. The security team has recommended using a bastion host to achieve this goal. What is the primary purpose of a bastion host in this scenario?
The primary purpose of a bastion host in the context of an air-gapped NKP environment is to serve as a secure gateway and deployment point for managing the air-gapped cluster without direct internet access.
Specific purposes of the bastion host:
How it enables air-gapped operations:
Without the bastion host, the air-gapped cluster would have no way to receive deployments, updates, or management commands while maintaining complete network isolation.
A Platform Engineer needs to do an air-gapped installation of NKP. This environment needs to run without Internet access and be fully operational, including updates. Docker has been installed, and the NKP bundle exists on a bastion host. What is the first command that the engineer must run to begin the process?
The first command that the engineer must run to begin the air-gapped NKP installation process is to load and prepare the NKP bundle. This typically involves running: docker load -i nkp-bundle.tar.gz to load the NKP container images from the bundle file into the local Docker registry on the bastion host. Alternatively, the engineer might run a bundle extraction or setup command such as tar -xzf nkp-bundle.tar.gz to extract the bundle contents. After loading the images, the next steps would be to tag and push them to the private registry, or configure the environment to use the locally loaded images for the air-gapped installation to proceed.
What prerequisite must the administrator ensure is met for this task?
Without seeing the specific task context, a common prerequisite for most NKP administrative tasks includes: ensuring the user has appropriate RBAC permissions (admin or operator role), verifying network connectivity to the cluster, confirming kubectl access to the management cluster, and ensuring any required prerequisites (such as backing up existing configurations, disabling admission webhooks if needed, or pre-staging resources) are completed. The specific prerequisite depends on the particular administrative task being performed.
A Platform Engineer wants to deploy a custom OS image for multiple NKP clusters for Nutanix AHV and AWS. Which two tools come bundled to facilitate creating and placing a custom image into the respective image repository?
NKP comes bundled with two key tools for custom OS image creation and deployment: Image Builder (used to create custom OS images) and Cluster API (used to manage the placement and provisioning of these images into the respective infrastructure providers). Image Builder handles the creation and customization of OS images, while Cluster API orchestrates their deployment across different platforms like Nutanix AHV and AWS, placing them into the appropriate image repositories for each infrastructure.
Recently, the reliability of some developer platforms has been impacted by out-of-memory errors, causing pods to crash and then remain in a pending state. A Cloud Engineer has been tasked with collecting and reviewing centralized metrics across the various platform environments to identify differences between them and better understand what is causing these performance issues. What should the engineer do to address this request?
To collect and review centralized metrics across platform environments and identify the causes of out-of-memory errors, the Cloud Engineer should: (1) Access the centralized monitoring and metrics platform provided by NKP (typically Prometheus and Grafana from kube-prometheus-stack); (2) Review memory utilization metrics across all platform clusters to identify patterns and differences; (3) Examine pod memory requests, limits, and actual consumption using Prometheus queries; (4) Use Grafana dashboards to visualize memory trends and identify which workloads or namespaces are consuming excessive memory; (5) Review metrics over time to understand memory growth patterns; (6) Correlate out-of-memory events with specific workloads, deployments, or configuration changes; (7) Use kube-state-metrics to understand pod eviction patterns and container restart events. This centralized metrics approach allows comparison across environments and helps identify configuration differences (such as memory limits or resource allocation) that may be causing issues in some environments but not others.
Refer to the exhibits.
A Cloud Administrator had provisioned a Kubernetes cluster named demo that is no longer actively being used. A Quick review from the Systems Engineer confirms that the following VMs are part of the demo Kubernetes cluster
How should the demo cluster be properly deleted?
To properly delete the demo Kubernetes cluster, the proper procedure is: (1) Use the NKP management interface (UI or CLI) to initiate cluster deletion; (2) If using the NKP UI, navigate to the cluster and select the delete or remove cluster option; (3) If using the CLI, use a command such as nkp cluster delete demo; (4) Monitor the deletion process to ensure all cluster resources, including the VMs identified by the Systems Engineer, are properly removed; (5) Verify that all associated infrastructure (VMs, persistent volumes, load balancers) are cleaned up; (6) If the cluster deletion becomes stuck, use the force delete option with appropriate caution; (7) Confirm in the infrastructure management tool (vSphere, AWS, Nutanix, etc.) that all demo cluster VMs have been terminated. Using NKP's deletion functionality ensures that all cluster-related resources are properly cleaned up, rather than manually deleting VMs which could leave orphaned resources or cause cluster inconsistency.
Exam domains verified against: Official Nutanix NCP-CN exam guide, last checked October 2026.
Set up the infrastructure foundation for NKP by seeding a private registry, creating a bootstrap cluster, choosing appropriate license tiers, configuring a bastion host, building machine images or preparing nodes, and gathering provider-specific information needed for successful cluster deployment.
Deploy and customize NKP clusters using Cluster API and configure Kommander for multi-cluster management. Apply the correct license tiers to match your deployment requirements and operational needs.
Manage production NKP clusters through authentication and authorization with Dex and OIDC, implement logging with Loki, configure backup and recovery with Velero, monitor health with Prometheus, enable cluster autoscaling, and execute lifecycle management tasks.
Sample question from this domain above: Q5
Organize multiple clusters using workspaces and projects, deploy and attach new clusters to workspaces, manage cluster lifecycle including detachment and deletion, and configure platform applications across your Kubernetes environment.
Common questions about the exam itself
