Use ReplicaSet and Deployment

Orchestrating applications with ReplicaSet and Deployment mechanisms

Use ReplicaSet to start application

Check list of namespaces:

kubectl get ns

Create new myapp namespace for Pod instance:

kubectl create ns myapp

Create a ReplicaSet template manifest for myapp:

rs-myapp.yaml

apiVersion: apps/v1
kind: ReplicaSet
metadata:
  name: myapp
  namespace: myapp
spec:
  selector:
    matchLabels:
      app: myapp
  replicas: 1
  template:
    metadata:
      labels:
        app: myapp
    spec:
      containers:
      - name: myapp
        image: ghcr.io/go4clouds/myapp:v1.0
        ports:
        - name: http
          containerPort: 8081
          protocol: TCP

Apply ReplicaSet manifest:

kubectl apply -f rs-myapp.yaml

Check ReplicaSet and Pod status:

kubectl get replicaset,pods -n myapp -o wide

Scale up the number of replicas:

kubectl scale -n myapp replicaset myapp --replicas=3

Check ReplicaSet and Pod status:

kubectl get rs,pods -n myapp -o wide

To clean up, remove the entire namespace:

kubectl delete ns myapp

Use Deployment to start application

Check list of namespaces:

kubectl get ns

Create a new myapp namespace for the Pod instance:

kubectl create ns myapp

Create a Deployment template manifest for 'myapp':

deploy-myapp.yaml

apiVersion: apps/v1
kind: Deployment
metadata:
  name: myapp
  namespace: myapp
spec:
  selector:
    matchLabels:
      app: myapp
  replicas: 3
  template:
    metadata:
      labels:
        app: myapp
    spec:
      containers:
      - name: myapp
        image: ghcr.io/go4clouds/myapp:v1.0
        ports:
        - name: http
          containerPort: 8081
          protocol: TCP

Apply the Deployment manifest:

kubectl apply -f deploy-myapp.yaml

Check Deployment, ReplicaSet and Pod status:

kubectl get events --sort-by='.lastTimestamp' -n myapp
kubectl get deploy,rs,pods -n myapp -o wide

Scale up the number of replicas:

kubectl scale -n myapp deployment myapp --replicas=5

Check the status of Deployment, ReplicaSet, and Pod:

kubectl get events --sort-by='.lastTimestamp' -n myapp
kubectl get deploy,rs,pods -n myapp -o wide

To clean up, remove the myapp namespace:

kubectl delete ns myapp

Testing rolling update deployment

Create webapp Deployment:

kubectl create namespace webapp
kubectl create deployment -n webapp webapp --image=docker.io/library/nginx:1.16.1

See `deploy-webapp.yaml`

apiVersion: apps/v1
kind: Deployment
metadata:
  creationTimestamp: null
  labels:
    app: webapp
  name: webapp
  namespace: webapp
spec:
  replicas: 1
  selector:
    matchLabels:
      app: webapp
  strategy: {}
  template:
    metadata:
      creationTimestamp: null
      labels:
        app: webapp
    spec:
      containers:
      - image: docker.io/library/nginx:1.16.1
        name: nginx
        resources: {}
status: {}

Check the status of Deployment:

kubectl get deploy,rs,pods -n webapp
kubectl rollout history deployment -n webapp webapp

Change number of replicas:

kubectl scale deployment -n webapp webapp --replicas=3

Annotate changes:

kubectl annotate deployment -n webapp webapp kubernetes.io/change-cause="scale up from 1 to 3 replicas"

Check the status of Deployment:

kubectl get deploy,rs,pods -n webapp

Update Deployment configuration:

export CONTAINER_NAME=$(kubectl get deployment -n webapp webapp -o jsonpath="{...name}" | cut -d " " -f2)
kubectl set image deployment -n webapp webapp $CONTAINER_NAME=docker.io/library/nginx:1.20.2

Annotate changes:

kubectl annotate deployment -n webapp webapp kubernetes.io/change-cause="$CONTAINER_NAME=docker.io/library/nginx:1.20.2"

Check the status of Deployment:

kubectl get deploy,rs,pods -n webapp

Check the history of Deployment rollout:

kubectl rollout history --revision=2 -n webapp deployment webapp

Change resource limits:

kubectl set resources -n webapp deployment webapp --containers=$CONTAINER_NAME --limits=cpu=200m,memory=512Mi

Annotate changes:

kubectl annotate deployment -n webapp webapp kubernetes.io/change-cause="changed limits to 200m for cpu and 512Mi for memory"

Try to break the Deployment and use a broken container image:

kubectl rollout history deployment -n webapp webapp
kubectl set image deployment -n webapp webapp $CONTAINER_NAME=docker.io/library/nginx:1.24.3-NON-EXISTING

Annotate changes:

kubectl annotate deployment -n webapp webapp kubernetes.io/change-cause="changed Nginx to a non-existing version"

Check the status of Deployment:

kubectl get deploy,rs,pods -n webapp
kubectl get events --sort-by='.lastTimestamp' -n webapp

Roll back changes for the Deployment because the previous update is broken due to the non-existing image nginx:1.24.3-NON-EXISTING:

kubectl rollout undo deployment -n webapp webapp 

Enable autoscaling

As a continuation of the previous step, we can enable HorizontalPodAutoscaler.

Enable Horizontal Pod Autoscaler for webapp deployment:

kubectl autoscale deployment -n webapp webapp --min=1 --max=5 --cpu-percent=60

See `autoscale.yaml`

apiVersion: autoscaling/v1
kind: HorizontalPodAutoscaler
metadata:
  creationTimestamp: null
  name: webapp
spec:
  maxReplicas: 5
  minReplicas: 1
  scaleTargetRef:
    apiVersion: apps/v1
    kind: Deployment
    name: webapp
  targetCPUUtilizationPercentage: 60
status:
  currentReplicas: 0
  desiredReplicas: 0

Check the status of HPA:

kubectl get hpa webapp -n webapp

Output:

NAME    REFERENCE          TARGETS   MINPODS   MAXPODS   REPLICAS   AGE
nginx   Deployment/nginx   0%/60%    1         5         1          36m

Check HPA event logs:

kubectl describe hpa -n webapp webapp

To clean up:

kubectl delete namespace webapp