Kubernetes architecture (control plane and nodes) - Practice Problems & Coding Challenges

Challenge - 5 Problems

🧠 Conceptual

intermediate

Control Plane Components Role

Which component of the Kubernetes control plane is responsible for maintaining the desired state of the cluster by watching the cluster state and making changes?

A kube-controller-manager

B kube-apiserver

C kube-scheduler

D etcd

💡 Hint

Think about the component that manages controllers to reconcile cluster state.

The kube-controller-manager runs controllers that watch the cluster state and make changes to reach the desired state.

💻 Command Output

intermediate

Node Status Command Output

What is the output of the command kubectl get nodes when the cluster has two nodes named node1 and node2, both ready?

Kubernetes

kubectl get nodes

A

node1   Ready    master   10d
node2   Ready    worker   10d

B

NAME    STATUS   ROLES    AGE
node1   Ready    <none>   10d
node2   Ready    <none>   10d

C

NAME    STATUS    ROLES    AGE
node1   NotReady  <none>   10d
node2   Ready    <none>   10d

D

NAME    STATUS   ROLES    AGE
node1   Ready    <none>   5m
node2   Ready    <none>   5m

💡 Hint

Check the exact column headers and typical output format.

The kubectl get nodes command lists nodes with columns NAME, STATUS, ROLES, and AGE. Both nodes are ready with no roles assigned.

🔀 Workflow

advanced

Pod Scheduling Process

Put the steps of the Kubernetes pod scheduling process in the correct order.

A 1,2,3,4

B 1,3,2,4

C 2,1,3,4

D 1,2,4,3

💡 Hint

Think about the logical flow from detecting a pod to assigning it to a node.

The scheduler first detects unscheduled pods, filters nodes, ranks them, then binds the pod to the best node.

❓ Troubleshoot

advanced

Diagnosing Node Not Ready State

You notice a node in your Kubernetes cluster is in NotReady state. Which of the following is the most likely cause?

A The etcd database is full and cannot store new data.

B The pod running on the node has crashed.

C The kube-scheduler is overloaded and cannot assign pods.

D The kubelet on the node is not running or has lost connection to the control plane.

💡 Hint

Consider what controls node health reporting.

The kubelet reports node status to the control plane. If it stops or loses connection, the node shows as NotReady.

✅ Best Practice

expert

High Availability of Control Plane

Which setup ensures high availability of the Kubernetes control plane?

A Run multiple etcd instances on worker nodes only.

B Run a single kube-apiserver with multiple kube-scheduler instances on the same node.

C Run multiple kube-apiserver instances behind a load balancer with replicated etcd cluster.

D Run multiple kubelet processes on the control plane node.

💡 Hint

Think about redundancy and fault tolerance for control plane components.

High availability requires multiple API servers behind a load balancer and a replicated etcd cluster to avoid single points of failure.