Knowledge: Kubernetes part3

Kubernetes part3

Class 87th Kubernetes Part3 August 13th

Multi node cluster

Cluster (group of master and worker server called cluster)

Operator server/User server (kubectl) --> send request to create pod-->API Server --> Scheduler decided

which worker node create and inform back to API server --> API server will create Pod in worker node.

Practical :

Step1: Create Ec2 instance Kops Instanace name take instance t3.micro okay, 20 GB disk for master c7.flex.large required,worker nodes 2 t3.micro

Install Kubectl https://kubernetes.io/docs/tasks/tools/

[root@ip-10-0-0-22 bin]# curl -LO "https://dl.k8s.io/release/$(curl -L -s https://dl.k8s.io/release/stable.txt)/bin/linux/amd64/kubectl"

% Total % Received % Xferd Average Speed Time Time Time Current

Dload Upload Total Spent Left Speed

100 138 100 138 0 0 1103 0 --:--:-- --:--:-- --:--:-- 1112

100 57.3M 100 57.3M 0 0 101M 0 --:--:-- --:--:-- --:--:-- 101M

[root@ip-10-0-0-22 bin]# mv kubectl /usr/local/bin

[root@ip-10-0-0-22 bin]# kubectl version

[root@ip-10-0-2-52 bin]# chmod 777 kubectl

[root@ip-10-0-2-52 bin]# kubectl version

Client Version: v1.33.4

Kustomize Version: v5.6.0

Step2: kops installation steps Linux directly past to the server

https://kops.sigs.k8s.io/getting_started/install/

[root@ip-10-0-0-22 ~]# curl -Lo kops https://github.com/kubernetes/kops/releases/download/$(curl -s https://api.github.com/repos/kubernetes/kops/releases/latest | grep tag_name | cut -d '"' -f 4)/kops-linux-amd64

[root@ip-10-0-0-22 ~]# chmod +x kops

[root@ip-10-0-0-22 ~]#sudo mv kops /usr/local/bin/kops

% Total % Received % Xferd Average Speed Time Time Time Current

Dload Upload Total Spent Left Speed

0 0 0 0 0 0 0 0 --:--:-- --:--:-- --:--:-- 0

100 274M 100 274M 0 0 48.3M 0 0:00:05 0:00:05 --:--:-- 52.1M

[root@ip-10-0-0-22 ~]# kops version

Client version: 1.33.0 (git-v1.33.0)

Step3: Cluster information ,storing in AWS S3 bucket

[root@ip-10-0-0-22 ~]# aws s3 ls

Unable to locate credentials. You can configure credentials by running "aws configure".

Attach the admin role to the Ec2 machine

Step4: Create one bucket version should be enabled "ccitpublicbucket16"

[root@ip-10-0-0-22 ~]# aws s3 ls

2025-08-16 16:58:08 ccitpublicbucket16

We need to inform kops cluster store the specific bucket store the cluster details

[root@ip-10-0-0-22 ~]# export KOPS_STATE_STORE=s3://ccitpublicbucket16

[root@ip-10-0-0-22 ~]#

Step5: Cluster creation

https://kops.sigs.k8s.io/getting_started/aws/

Syntax:-

kops create cluster \ --name=${NAME} \ --cloud=aws \ --zones=us-west-2a \ --discovery-store=s3://prefix-example-com-oidc-store/${NAME}/discovery

Cluster name should be <anyname>.k8s.local

Below command ran once it will take (master and woker node will create at the end give some suggestion keep for same)

[root@ip-10-0-0-22 ~]# kops create cluster --name kopsclstr.k8s.local --zones eu-west-2b,eu-west-2a --master-count 1 --master-size c7i-flex.large --master-volume-size 20 --node-count 2 --node-size t3.micro --node-volume-size=15 --image=ami-044415bb13eee2391

Suggestions:

* list clusters with: kops get cluster

* edit this cluster with: kops edit cluster kopsclstr.k8s.local

* edit your node instance group: kops edit ig --name=kopsclstr.k8s.local nodes-eu-west-2b

* edit your control-plane instance group: kops edit ig --name=kopsclstr.k8s.local control-plane-eu-west-2b

Finally configure your cluster with: kops update cluster --name kopsclstr.k8s.local --yes --admin

Step6: as given suggestion run the command , it will take time to create

[root@ip-10-0-0-22 ~]# kops update cluster --name kopsclstr.k8s.local --yes --admin

kOps has set your kubectl context to kopsclstr.k8s.local

Cluster is starting. It should be ready in a few minutes.

Suggestions:

* validate cluster: kops validate cluster --wait 10m

* list nodes: kubectl get nodes --show-labels

* ssh to a control-plane node: ssh -i ~/.ssh/id_rsa ubuntu@

* the ubuntu user is specific to Ubuntu. If not using Ubuntu please use the appropriate user based on your OS.

* read about installing addons at: https://kops.sigs.k8s.io/addons.

Successfully created control plane node and two worker nodes

Target Groups

Load balancer

Autoscaling

VPC

In real time we are set infra using terraform script to eks -- it will take 45 minute

But Kops it will completed in -5 minutes

Project we will do using terraform

Step7: it will take some time minimum 10 mints to 15 minutes

[root@ip-10-0-0-22 ~]# kops validate cluster --wait 10m

[[root@ip-10-0-0-22 ~]# kubectl get nodes

NAME STATUS ROLES AGE VERSION

i-03d052cacf647b877 Ready node 59s v1.32.4

i-0e8a3f4c3da2f1696 Ready node 61s v1.32.4

i-0f5f236564f4f7804 Ready control-plane 3m11s v1.32.4

Step8: We have successfully deploy the pods

I Can’t able to access my application

The reason why we are not able access the application : In Kubernetes, if you want to access the application we have to expose our pod. To expose these pods we use Kubernetes services.

Through Services in Kubernetes ,We export the pods different types

Services: Cluster-Ip,node port,load balancer , external dns(not using)

Every pod have default one ip,with that ip we cannot able access application,we expose that ip

1. Cluster-IP : using this cluster-ip service we can export pod, it use only for internal scop internal,external we can able access the application internet

For ex:- it will provide stable ip or static ip using internal only ,use for database expose pod

using curl, we can test

2.Node Port : it will give you Stable ip and also give the range node port range(30k-32656)

with this port range number , using worker node public ip access the application in the pod

Step8: Se below master node or control-plane and node or worker node are ready

Server information :kubectl get no -o also work

[root@ip-10-0-0-22 ~]# kubectl get node -o wide

NAME STATUS ROLES AGE VERSION INTERNAL-IP EXTERNAL-IP OS-IMAGE KERNEL-VERSION CONTAINER-RUNTIME

i-03d052cacf647b877 Ready node 10m v1.32.4 172.20.151.201 3.8.130.16 Ubuntu 24.04.2 LTS 6.8.0-1029-aws containerd://1.7.28

i-0e8a3f4c3da2f1696 Ready node 10m v1.32.4 172.20.72.207 3.10.152.129 Ubuntu 24.04.2 LTS 6.8.0-1029-aws containerd://1.7.28

i-0f5f236564f4f7804 Ready control-plane 12m v1.32.4 172.20.180.173 35.176.113.95 Ubuntu 24.04.2 LTS 6.8.0-1029-aws containerd://1.7.28

Step9: For pod creation manifest file need prepare

Docker installation

[root@ip-10-0-0-22 ~]# yum install docker -y && systemctl start docker

[root@ip-10-0-0-22 ~]# vi manifest.yaml

[root@ip-10-0-0-22 ~]# cat manifest.yaml

apiVersion: v1

kind: Pod

metadata:

name: pod-1

labels:

app: zamota

spec:

containers:

- name: container1

image: shaikmustafa/dm

ports:

- containerPort: 80

Step10: it is asking user name and password , i have create own docker image

[root@ip-10-0-0-22 ~]# kubectl create -f manifest.yaml

pod/pod-1 created

Created pod above successfully.

Step11: One pod created successfully

[root@ip-10-0-0-22 ~]# kubectl get po -o wide

NAME READY STATUS RESTARTS AGE IP NODE NOMINATED NODE READINESS GATES

pod-1 1/1 Running 0 63s 100.96.3.131 i-03d052cacf647b877 <none> <none>

Step12: How to check, pod details ,as you see scheduler tell ,create particular instance to create the pod

[root@ip-10-0-0-22 ~]# kubectl describe pod pod-1

Events:

Type Reason Age From Message

---- ------ ---- ---- -------

Normal Scheduled 3m15s default-scheduler Successfully assigned default/pod-1 to i-03d052cacf647b877

Normal Pulling 3m14s kubelet Pulling image "shaikmustafa/dm"

Normal Pulled 3m9s kubelet Successfully pulled image "shaikmustafa/dm" in 4.892s (4.892s including waiting). Image size: 60099640 bytes.

Normal Created 3m9s kubelet Created container: container1

Normal Started 3m9s kubelet Started container container1

Step13: Service need to create ,which service going to expose , go to that selector cluster

clusterip ,we have given po port 80,we have give same to give port to service port also 80

[root@ip-10-0-0-22 ~]# kubectl get po --show-labels

NAME READY STATUS RESTARTS AGE LABELS

pod-1 1/1 Running 0 53m app=zamoto

[root@ip-10-0-0-22 ~]# vim service.yaml

[root@ip-10-0-0-22 ~]# vi service.yaml

---

apiVersion: v1

kind: Service

metadata:

name: myservice

spec:

type: ClusterIP

selector:

app: zamato

ports:

- port: 80

targetPort: 80

[root@ip-10-0-0-22 ~]# kubectl create -f service.yaml

service/myservice created

Service create kubernetes default one , mysevice we have create our own

[root@ip-10-0-0-22 ~]# kubectl get svc

NAME TYPE CLUSTER-IP EXTERNAL-IP PORT(S) AGE

kubernetes ClusterIP 100.64.0.1 <none> 443/TCP 23m

myservice ClusterIP 100.68.51.206 <none> 80/TCP 10s

Cluster Stable ip 100.68.51.206 we can able access internal only using work/slave public ip

Step14: Worker node

[root@ip-10-0-0-22 ~]# kubectl label pod pod-1 app=zamato --overwrite

pod/pod-1 labeled

[root@ip-10-0-0-22 ~]# kubectl get endpoints myservice

NAME ENDPOINTS AGE

myservice 100.96.3.131:80 21m

Internal working worker/slave node stable link working fine

Step15: service Command

[root@ip-10-0-0-22 ~]# kubectl get svc

NAME TYPE CLUSTER-IP EXTERNAL-IP PORT(S) AGE

kubernetes ClusterIP 100.64.0.1 <none> 443/TCP 58m

myservice ClusterIP 100.68.51.206 <none> 80/TCP 35m

Full information of the service

[root@ip-10-0-0-22 ~]# kubectl get svc -o wide

NAME TYPE CLUSTER-IP EXTERNAL-IP PORT(S) AGE SELECTOR

kubernetes ClusterIP 100.64.0.1 <none> 443/TCP 59m <none>

myservice ClusterIP 100.68.51.206 <none> 80/TCP 36m app=zamato

[root@ip-10-0-0-22 ~]#

if you need yaml format

[root@ip-10-0-0-22 ~]# kubectl get svc -o yaml

apiVersion: v1

Json

[root@ip-10-0-0-22 ~]# kubectl get svc -o json

Full information one service

[root@ip-10-0-0-22 ~]# kubectl describe svc myservice

Name: myservice

Namespace: default

Labels: <none>

Annotations: <none>

Selector: app=zamato

Type: ClusterIP

IP Family Policy: SingleStack

IP Families: IPv4

IP: 100.68.51.206

IPs: 100.68.51.206

Port: <unset> 80/TCP

TargetPort: 80/TCP

Endpoints: 100.96.3.131:80

Session Affinity: None

Internal Traffic Policy: Cluster

Events: <none>

Delete the service

[root@ip-10-0-0-22 ~]# kubectl delete svc myservice

service "myservice" deleted

[root@ip-10-0-0-22 ~]# kubectl get svc

NAME TYPE CLUSTER-IP EXTERNAL-IP PORT(S) AGE

kubernetes ClusterIP 100.64.0.1 <none> 443/TCP 63m

Node Port

It will provide you

stable ip --Internal use

Node Port --Internal and external through internet (slave server using public ip along with node port -IP)

node port range : 30k -32656

Step1: Existing port just change the cluster IP to node Ip in servie.yaml file

[root@ip-10-0-0-22 ~]# vi service.yaml

[root@ip-10-0-0-22 ~]# cat service.yaml

---

apiVersion: v1

kind: Service

metadata:

name: myservice

spec:

type: NodePort

selector:

app: zamato

ports:

- port: 80

targetPort: 80

[root@ip-10-0-0-22 ~]# kubectl create -f service.yaml

service/myservice created

[root@ip-10-0-0-22 ~]# kubectl get svc

NAME TYPE CLUSTER-IP EXTERNAL-IP PORT(S) AGE

kubernetes ClusterIP 100.64.0.1 <none> 443/TCP 69m

myservice NodePort 100.71.48.35 <none> 80:31048/TCP 21s

Step2:

Stable ip try access internally 100.71.48.35

Step3: worker/slave node public ip

curl 3.10.152.129:31048

Step4:

Try to open internet browser , if not able access open the port in security group 31048

Worker node 1

Worker node 2 with public ip

Step5: Real time we are not use ip address, Load balancer service

[root@ip-10-0-0-22 ~]# kubectl delete svc myservice

service "myservice" deleted

Load balance :

Work it work load balancer in kubernetes

K8s svc -->Cloude provide -->LB (DNS) name wil provide -- >cctter.com

It will give these use case

Stable ip

Node port

DNS

[root@ip-10-0-0-22 ~]# vi service.yaml

[root@ip-10-0-0-22 ~]# cat service.yaml

---

apiVersion: v1

kind: Service

metadata:

name: myservice

spec:

type: LoadBalancer

selector:

app: zamato

ports:

- port: 80

targetPort: 80

[root@ip-10-0-0-22 ~]# kubectl create -f service.yaml

service/myservice created

[root@ip-10-0-0-22 ~]# kubectl get svc

NAME TYPE CLUSTER-IP EXTERNAL-IP PORT(S) AGE

kubernetes ClusterIP 100.64.0.1 <none> 443/TCP 86m

myservice LoadBalancer 100.68.221.171 a7655f520a3d44cd98e3fdff6def1af1-351032056.eu-west-2.elb.amazonaws.com 80:32627/TCP 21s

Step6: checking with stable IP

ubuntu@i-03d052cacf647b877:~$ curl 100.68.221.171

<!doctype html>

Step7: Able to access application node port

Step8: Need to with loadbalancer/dns name , it will not work immediately ,it will take time to

1 minute to comunicate

a7655f520a3d44cd98e3fdff6def1af1-351032056.eu-west-2.elb.amazonaws.com

Now load will transfer to two workers not equally.

Step 9: Need to delete the Cluster after completion using command only not manually , if you are manually delete ,autoscaling trying create parallel ,so do command

Below delete the cluster give this command

[root@ip-10-0-0-22 ~]#export KOPS_STATE_STORE=s3://ccitpublicbucket16

Out cluster information will store in this s3 bucket ccitpublicbucket16

Step10:

API endpoint and services:

[root@ip-10-0-0-22 ~]# kubectl cluster-info

Kubernetes control plane is running at https://api-kopsclstr-k8s-local-0kba4s-37d4891bf709009e.elb.eu-west-2.amazonaws.com

CoreDNS is running at https://api-kopsclstr-k8s-local-0kba4s-37d4891bf709009e.elb.eu-west-2.amazonaws.com/api/v1/namespaces/kube-system/services/kube-dns:dns/proxy

To further debug and diagnose cluster problems, use 'kubectl cluster-info dump'.

cluster state and configuration:

[root@ip-10-0-0-22 ~]# kops get cluster

NAME CLOUD ZONES

kopsclstr.k8s.local aws eu-west-2a,eu-west-2

node health:

[root@ip-10-0-0-22 ~]# kubectl get nodes

NAME STATUS ROLES AGE VERSION

i-03d052cacf647b877 Ready node 103m v1.32.4

i-0e8a3f4c3da2f1696 Ready node 103m v1.32.4

i-0f5f236564f4f7804 Ready control-plane 105m v1.32.4

system components:

[root@ip-10-0-0-22 ~]# kubectl get pods -n kube-system

[root@ip-10-0-0-22 ~]# kops delete cluster --name kopsclstr.k8s.local --yes

subnet:subnet-014cdb83033135899 ok

security-group:sg-0c7c7e35635a07319 ok

route-table:rtb-0d6c56defcbad4db3 ok

vpc:vpc-08692f993cc46837f ok

dhcp-options:dopt-03b1723c1a058a166 ok

Deleted kubectl config for kopsclstr.k8s.local

Deleted cluster: "kopsclstr.k8s.local"

--All will create , Autoscaling,instance,loadbalancer,

KOPS Server only you need manually delete.

Reference document:

https://mustafa-k8s.hashnode.dev/kops-kubernetes-operations-the-ultimate-guide-for-devops-engineers

--Thanks

Knowledge

Wednesday, August 13, 2025

Kubernetes part3

No comments:

Post a Comment