04-controllers.md

There are multiple types of Controllers in Kubernetes

Replication Controller

Follow below commnads and observe how pods are getting created

kubectl create -f kubia-rc.yaml
kubectl get pods
kubectl delete pod kubia-53thy
kubectl get pods
kubectl get rc

Observe RC

kubectl describe rc kubia

Modify a pod's label

 kubectl get pods --show-labels
 kubectl label pod <podname> app=foo --overwrite

observe the pods with label column [new pod being created]

kubectl get pods -L app

Scale out

kubectl scale rc kubia --replicas=10

Delete without deleting pods

kubectl delete rc kubia --cascade=false

Replica set

Replica Set creation

kubectl create -f kubia-replicaset.yaml

kubectl get rs 
kubectl describe rs

Daemon set

This daemon set is designed to run on all the disks that has ssd disk (disk=ssd )

  kubectl create -f ssd-monitor-daemonset.yaml
  kubectl get ds
  kubectl get po
  kubectl get node
  kubectl label node <nodename> disk=ssd

check it now

 kubectl get po

job

  kubectl get jobs
  kubectl get po

After the two minutes have passed, see the status "completed"

 kubectl get po -a
 kubectl logs <jobpodname>

Sequential completion and parallelism

kubectl create -f multi-completion-batch-job.yaml
kubectl create -f multi-completion-parallel-batch-job.yaml

You can even change a Job’s parallelism property while the Job is running

kubectl scale job multi-completion-batch-job --replicas 3