CodeBlanks Logo Icon CodeBlanks
Login Register

Worker node

Learn kubelet and kube-proxy on worker nodes.

The worker node

Three components run on each worker node:

  • kubelet: takes the pods assigned to it from the API server and starts them with the container runtime; reports the pods' status back.
  • Container runtime: actually runs the containers (containerd, CRI-O).
  • kube-proxy: manages the network rules that route traffic arriving at Service addresses to the right pods.
$ kubectl describe node worker-1 | grep -A3 "System Info"
 Container Runtime Version:  containerd://1.7.22
 Kubelet Version:            v1.31.2

Node capacity

Each node has CPU and memory capacity; the scheduler places pods according to this capacity. kubectl describe node shows the allocatable resources and the total requests of the pods on the node; if metrics-server is installed, kubectl top gives the current usage:

$ kubectl top nodes
NAME       CPU(cores)   CPU(%)   MEMORY(bytes)   MEMORY(%)
worker-1   312m         15%      1842Mi          47%
worker-2   188m         9%       1203Mi          31%

Practice Inspect a node

Write two commands: show the details of the worker-1 node; then show the current CPU and memory usage of all nodes.

Starter code

terminal
# buraya yaz

Sample solution

terminal
kubectl describe node worker-1
kubectl top nodes

Expected output

Name:               worker-1
Roles:              <none>
Allocatable:
  cpu:                2
  memory:             3912Mi
NAME       CPU(cores)   CPU(%)   MEMORY(bytes)   MEMORY(%)
worker-1   312m         15%      1842Mi          47%

Questions and answers

Review the lesson questions together with their answers. To test yourself, solve the lesson interactively.

1 Does Kubernetes need Docker to run containers?

  • Only on Windows nodes
  • No; a CRI-compatible runtime (e.g. containerd) is enough Correct answer
  • Yes, always
  • Only on the control plane

Explanation: Images built with Docker follow the OCI standard and run with containerd.

2 Show the nodes' current resource usage.

Question 
kubectl ____ nodes
Answer 
kubectl top nodes

Explanation: kubectl top reads metrics-server data.

3 Match.

Term Meaning
kubelet The agent on the node
containerd Container runtime
kube-proxy Service routing
312m 0.312 CPU cores

All lessons in Intermediate

1 What is Kubernetes?

  1. What is Kubernetes?
  2. Why Kubernetes?
  3. Cluster architecture
  4. Control plane
  5. Worker node
  6. Summary - What is Kubernetes

2 Minikube

  1. What is Minikube?
  2. Installing Minikube
  3. minikube start
  4. minikube status
  5. minikube dashboard
  6. Summary - Minikube

3 kubectl basics

  1. What is kubectl?
  2. kubectl get
  3. kubectl describe
  4. kubectl apply
  5. kubectl delete
  6. Summary - kubectl

4 Pod

  1. What is a Pod?
  2. Pod YAML
  3. Creating a Pod
  4. Pod status
  5. Multi-container Pod
  6. Summary - Pod

5 Deployment

  1. What is a Deployment?
  2. Deployment YAML
  3. Replica count
  4. Rolling update
  5. Rollback
  6. Summary - Deployment

6 ReplicaSet

  1. What is a ReplicaSet?
  2. ReplicaSet YAML
  3. ReplicaSet selector
  4. Scaling
  5. Self-healing
  6. Summary - ReplicaSet

7 Service

  1. What is a Service?
  2. ClusterIP
  3. NodePort
  4. LoadBalancer
  5. Service selector
  6. Summary - Service

8 ConfigMap

  1. What is a ConfigMap?
  2. ConfigMap YAML
  3. Environment variable
  4. Mounting as a volume
  5. kubectl create configmap
  6. Summary - ConfigMap

9 Secret

  1. What is a Secret?
  2. Secret types
  3. Secret YAML
  4. Mounting a Secret
  5. kubectl create secret
  6. Summary - Secret

10 Namespace

  1. What is a Namespace?
  2. kubectl get namespaces
  3. Creating a Namespace
  4. The -n flag
  5. Default Namespace
  6. Summary - Namespace

11 Labels and selectors

  1. What is a label?
  2. Adding labels
  3. Label selector
  4. kubectl get -l
  5. Equality and set selectors
  6. Summary - Labels

12 Project: Multi-tier app

  1. Project design
  2. Database Deployment
  3. Web Deployment
  4. ConfigMap and Secret
  5. Putting it together - multi-tier

13 Troubleshooting

  1. kubectl logs
  2. kubectl exec
  3. kubectl describe pod
  4. CrashLoopBackOff
  5. Pending Pod
  6. Summary - Troubleshooting

14 Intermediate final

  1. Challenge - Pod YAML
  2. Challenge - Deployment
  3. Challenge - Service
  4. Challenge - ConfigMap
  5. Challenge - Intermediate project