connect to existing Cluster

To update existing kubectl cluster To local kubeconfig For Connecting

  aws eks update-kubeconfig --name <cluster-name> --profile 
  ```
  ```bash
  aws eks update-kubeconfig --name seeding --profile seeding --region us-east-1

To Get Service

  kubectl get svc

  eksctl create cluster --name selenium --region ap-south-1
  ```
  ```bash
  eksctl create cluster "selenium" --zones us-east-1a,us-east-1b,us-east-1c,us-east-1d,us-east-1f
  ```
  ```bash
  eksctl register cluster --name selenium --provider other --region us-east-1
  ```
```bash
  aws eks update-kubeconfig --region us-east-1 --name selenium

To delete existing kubectl config

  rm ~/.kube/config

  helm install selenium-grid docker-selenium/selenium-grid
  kubectl get all -n selenium-grid
  kubectl get services

To connect to internal network using busybox

  kubectl run -i --tty busybox --image=busybox --restart=Never -- sh

For centos pod

  kubectl run -i --tty centos --image=centos --restart=Never -- sh

To get all resource details

  kubectl get all

To delete the deployment

  kubectl delete deployments -l app=selenium-edge-node

To Describe Events In Running Pods

  kubectl describe pods -l app=selenium-chrome-node | grep Events -A4

To Force Delete A Pod

  kubectl delete pod selenium-chrome-node --grace-period=0 --force
  kubectl delete pod selenium-edge-node --grace-period=0 --force
  kubectl delete pod selenium-firefox-node --grace-period=0 --force
  kubectl delete pod selenium-hub-c6c94c6c4-h558k --grace-period=0 --force

To get status of kube system

  kubectl get pods -n=kube-system | grep coredns

To scale pod replicaset

  kubectl scale --replicas=5 replicaset/selenium-chrome-node-7bf4f8dc77
  kubectl scale deployment.apps/selenium-node-chrome-node --replicas=1

scale deployment

  kubectl scale deployment/selenium-chrome-node --replicas=2

To get the value replicaset

  kubectl get rs selenium-chrome-node-5f44bffc9b -o jsonpath="{.status.replicas} {.status.availableReplicas}"

To get replica status

  kubectl get rs

Run the following command to forward a local port to the service:

  kubectl port-forward service/selenium-hub 4444:4444

To get running resource to yaml

  crd = CustomResourceDefinition 
  rs  = replicaset
  kubectl get crd <CRD-NAME> -o yaml

If CRD unable to Delete remove the crd using kubectl edit

link-to-delete-crd[https://learn.microsoft.com/en-us/answers/questions/602466/custom-crds-not-getting-deleted-in-aks-cluster-how]

  kubectl edit crd crd-name   #  Remove finalizers from it and save

eks cluster with command

working with fargate,image is downloading & running

 eksctl create cluster --name seeding --region us-east-1 --version 1.26 --fargate --profile

seeding Next: allow 80 port in security group for loadbalancer access

To get metric server

  kubectl get --raw "/apis/external.metrics.k8s.io/v1beta1"

To Find which Namespce if metric in

  kubectl get scaledobject selenium-chrome-scaledobject -n default -o jsonpath={.status.externalMetricNames}

describe scaledobject

  kubectl describe scaledobject

To create node group

  eksctl create nodegroup --cluster=seeding \
  --name=node2 \
  --node-type=c5.2xlarge \
  --nodes=3 \
  --nodes-min=3 \
  --nodes-max=3 \
  --node-volume-size=20 \
  --ssh-access \
  --ssh-public-key=6548652153 \
  --managed \
  --region ap-south-1 
  ```


### To delete node group
```bash
  eksctl delete nodegroup --cluster=seeding --region ap-south-1
  eksctl delete --cluster=seeding --region ap-south-1

To Proxypass service

  kubectl --namespace monitoring port-forward svc/prometheus-k8s 9090

  kubectl kubernetes port-forward service/kubernetes 9090

To connect to ssh to pod

  kubectl exec --stdin --tty busybox -- /bin/bash
  kubectl exec --it  busybox -n namespace -- /bin/bash

Field selector

To delete all pods

Getting Specific Evicted is not working

kubectl delete pods --field-selector status.phase=Failed

kubectl logs -n {namespace}  (podname) --tail=100

If above not working using --previous=false

kubectl logs -p dbc-ddl-service-ddc4cf6f9-6dqx8 -n {namespace} --previous=false --tail=50

kubectl logs -p (podname)  --since=30m --timestamps 

k get events -n {namespace} --field-selector involvedobject.name={pod_name} --sort-by='.metadata.creationTimestamp'

k describe pod {pod_name} -n {namespace}

Get the count of the pods

kubectl get pods -n {namespace} | grep "dbc-ddl-service" | wc -l

scale replica

kubectl scale  deployment mysql --replicas=3

Delete pods where status is Evicted

kubectl delete pod $(kubectl get pods  --field-selector=status.phase=Failed -o jsonpath='{.items[?(@.status.reason=="Evicted")].metadata.name}')

rollout deployment

kubectl rollout restart deployment {podname} -n lifion

get_the status of container_restart count

kubectl get pods --sort-by='.status.containerStatuses[0].restartCount'

---

Related Pages

• [[basics]]
• [[CheatSheet]]
• [[Tips]]
• [[AWS/EC2]]
• [[kube-errors/conf.refuse]]