Understand Extension Interfaces (CNI, CSI, CRI, Etc.) CKA Practice Questions (Page 5)

Questions 21–25

1. 21foundation · easy

   What is the common benefit of extension interfaces like CNI, CSI, and CRI in Kubernetes?

   • AThey replace the need for kubelet and kube-proxy
   • BThey are required for all Kubernetes clusters to function
   • CThey allow Kubernetes to work with different network, storage, and runtime implementations without code changes
   • DThey provide a single unified implementation for all cluster features
   Reveal Solution Discussion Select an answer first
2. 22expert · hard

   A cluster uses a CNI plugin that does not support NetworkPolicy. The security team requires NetworkPolicy enforcement. The team wants to avoid a full cluster reinstall and minimize disruption. What is the best approach?

   • ACreate NetworkPolicy objects and rely on the API server to enforce them
   • BUse a service mesh to enforce network policies at the application layer
   • CEnable NetworkPolicy in the kubelet configuration
   • DInstall a second CNI plugin that supports NetworkPolicy and configure it to take over pod networking
   Reveal Solution Discussion Select an answer first
4. 23application · medium

   A team is using a CSI driver that supports snapshots. They want to create a new PVC from an existing snapshot. Which Kubernetes resource must be created first?

   • AVolumeSnapshotClass
   • BVolumeSnapshot
   • CStorageClass
   • DPersistentVolume
   Reveal Solution Discussion Select an answer first
5. 24foundation · easy

   Which component in a Kubernetes cluster typically acts as the CNI plugin executor?

   • Aetcd
   • Bkubelet
   • Ckube-proxy
   • Dkube-controller-manager
   Reveal Solution Discussion Select an answer first
6. 25application · medium

   A Kubernetes administrator needs to run a mix of standard containers and a few sandboxed workloads that require stronger isolation. The cluster currently uses containerd. The team wants to add a second runtime without recompiling the kubelet or changing how pods are scheduled. What should the administrator do?

   • AConfigure a RuntimeClass that references a CRI-compliant sandbox runtime and select it in the pod spec
   • BUse a mutating admission webhook to rewrite every pod to use the sandbox runtime
   • CReplace containerd with Docker and use Docker's native sandboxing
   • DInstall the new runtime on all nodes and set it as the default in the kubelet config
   Reveal Solution Discussion Select an answer first