NCP-CN practice question 15 of 25

[Conduct NKP Fleet Management] A Platform Engineer is deploying a new Kubernetes application in Amazon Web Services (AWS) Elastic Container Service…

Full set: 108 questions
[Conduct NKP Fleet Management] A Platform Engineer is deploying a new Kubernetes application in Amazon Web Services (AWS) Elastic Container Service for Kubernetes (EKS) and Azure Kubernetes Service (AKS). The engineer’s team has decided to use a custom image instead of the default images provided by AWS or Azure for their clusters. What functionality will the engineer lose in both AWS EKS and Azure AKS by choosing to use a custom image?
Answer options
A: Native monitoring tools, logging, and alerting
B: Cluster networking and load balancing
C: Built-in autoscaling and security capabilities Correct answer
D: Ability to use GPUs and persistent storage
Correct Answer:
C. Built-in autoscaling and security capabilities
Explanation

When using a custom image for worker nodes in managed Kubernetes services like AWS EKS or Azure AKS, you forgo the provider's optimizations. The default, provider-supplied images are pre-configured with essential components, including bootstrap scripts that enable nodes to automatically discover and join the cluster's control plane. This bootstrap process is critical for built-in autoscaling to function correctly; without it, newly scaled nodes will fail to register. Furthermore, these default images are hardened, patched, and maintained by the cloud provider, so using a custom image transfers the full responsibility for security configuration and ongoing maintenance to the user, thereby losing the built-in security capabilities.

Why Incorrect

A. Native monitoring tools, logging, and alerting: This functionality is not lost but requires manual setup. The engineer must install and configure the necessary monitoring and logging agents on the custom image.
B. Cluster networking and load balancing: These are core Kubernetes functions managed by the control plane and network plugins (CNI). They can be configured to work with custom node images.
D. Ability to use GPUs and persistent storage: The ability is not lost. Using these features with a custom image simply requires manual installation and configuration of the appropriate drivers (e.g., NVIDIA drivers, CSI drivers).

References

1. AWS EKS Documentation - Amazon EKS optimized AMIs: "The Amazon EKS optimized AMIs are configured to serve as the base image for Amazon EKS nodes... The AMI includes containerd

kubelet

the AWS IAM Authenticator

and the Amazon EKS bootstrap script." This bootstrap script is essential for node registration and

by extension

autoscaling.

URL: https://docs.aws.amazon.com/eks/latest/userguide/eks-optimized-ami.html

2. AWS EKS Documentation - Launching self-managed Amazon Linux nodes: This guide details the manual steps required when not using an EKS-optimized AMI

including configuring user data to run the bootstrap script (/etc/eks/bootstrap.sh)

demonstrating that this built-in automation is lost.

URL: https://docs.aws.amazon.com/eks/latest/userguide/launch-workers.html (See "Step 5: Configure node user data").

3. Azure AKS Documentation - Customize node configuration: "When you use a custom image

you're responsible for ensuring your image meets the AKS prerequisites... AKS doesn't provide support for custom images." This transfers all responsibility for security and compatibility

which is required for features like autoscaling

to the user.

URL: https://learn.microsoft.com/en-us/azure/aks/custom-node-configuration

Scroll to Top