Understanding and Combining GKE Autoscaling Strategies

“Understanding and Combining GKE Autoscaling Strategies”

Pengantar

Google Kubernetes Engine memiliki solusi horizontal dan vertikal untuk menskalakan pod dan infrastruktur Anda secara otomatis. Dalam hal pengoptimalan biaya, alat ini menjadi sangat berguna untuk memastikan bahwa beban kerja Anda dijalankan seefisien mungkin dan Anda hanya membayar apa yang Anda gunakan.

Praktikum

Task 1. Scale pods with Horizontal Pod Autoscaling

• Cek deployment

kubectl get deployment

• setting horizontal -autoscale deployment untuk php-apache

kubectl autoscale deployment php-apache --cpu-percent=50 --min=1 --max=10

• cek status scaling pod

kubectl get hpa

Task 2. Scale size of pods with Vertical Pod Autoscaling

• cek demo scaling

gcloud container clusters describe scaling-demo | grep ^verticalPodAutoscaling -A 1

• deploy helo-server

kubectl create deployment hello-server --image=gcr.io/google-samples/hello-app:1.0

• cek deployment

kubectl get deployment hello-server

• tambahkan cpu request 450 m untuk deployment

kubectl set resources deployment hello-server --requests=cpu=450m

• cek detail container

kubectl describe pod hello-server | sed -n "/Containers:$/,/Conditions:/p"

• Buat manifest Vertical Pod Autoscaler

cat << EOF > hello-vpa.yaml
apiVersion: autoscaling.k8s.io/v1
kind: VerticalPodAutoscaler
metadata:
  name: hello-server-vpa
spec:
  targetRef:
    apiVersion: "apps/v1"
    kind:       Deployment
    name:       hello-server
  updatePolicy:
    updateMode: "Off"
EOF

• apply manifest

kubectl apply -f hello-vpa.yaml

• cek detail

kubectl describe vpa hello-server-vpa

• update manifest

sed -i 's/Off/Auto/g' hello-vpa.yaml
kubectl apply -f hello-vpa.yaml

• buat scaling replica

kubectl scale deployment hello-server --replicas=2

• cek pod

kubectl get pods -w

Task 3. HPA results

• cek HPA

kubectl get hpa

Task 4. VPA results

• detail pod

kubectl describe pod hello-server | sed -n "/Containers:$/,/Conditions:/p"

Task 5. Cluster autoscaler

• enable autoscaling pada cluster

gcloud beta container clusters update scaling-demo --enable-autoscaling --min-nodes 1 --max-nodes 5

gcloud beta container clusters update scaling-demo \
--autoscaling-profile optimize-utilization

• buka Navigation menu, select Kubernetes Engine > Clusters
• pilih scaling-demo cluster
• cek system kube

kubectl get deployment -n kube-system

• Buat Pod Disruption Budget untuk semua kube-system

kubectl create poddisruptionbudget kube-dns-pdb --namespace=kube-system --selector k8s-app=kube-dns --max-unavailable 1
kubectl create poddisruptionbudget prometheus-pdb --namespace=kube-system --selector k8s-app=prometheus-to-sd --max-unavailable 1
kubectl create poddisruptionbudget kube-proxy-pdb --namespace=kube-system --selector component=kube-proxy --max-unavailable 1
kubectl create poddisruptionbudget metrics-agent-pdb --namespace=kube-system --selector k8s-app=gke-metrics-agent --max-unavailable 1
kubectl create poddisruptionbudget metrics-server-pdb --namespace=kube-system --selector k8s-app=metrics-server --max-unavailable 1
kubectl create poddisruptionbudget fluentd-pdb --namespace=kube-system --selector k8s-app=fluentd-gke --max-unavailable 1
kubectl create poddisruptionbudget backend-pdb --namespace=kube-system --selector k8s-app=glbc --max-unavailable 1
kubectl create poddisruptionbudget kube-dns-autoscaler-pdb --namespace=kube-system --selector k8s-app=kube-dns-autoscaler --max-unavailable 1
kubectl create poddisruptionbudget stackdriver-pdb --namespace=kube-system --selector app=stackdriver-metadata-agent --max-unavailable 1
kubectl create poddisruptionbudget event-pdb --namespace=kube-system --selector k8s-app=event-exporter --max-unavailable 1

• cek list node

kubectl get nodes

Task 6. Node Auto Provisioning

• Enable node auto provisioning

gcloud container clusters update scaling-demo \
    --enable-autoprovisioning \
    --min-cpu 1 \
    --min-memory 2 \
    --max-cpu 45 \
    --max-memory 160

Task 7. Test with larger demand

• buka terminal lain
• eksekusi perintah send query looping

kubectl run -i --tty load-generator --rm --image=busybox --restart=Never -- /bin/sh -c "while sleep 0.01; do wget -q -O- http://php-apache; done"

• cek hpa di terminal yg pertama

kubectl get hpa

• monitor cluster hendle

kubectl get deployment php-apache

Task 8. Optimize larger loads

• formula

• Buat manifest untuk pod yang di pause

cat << EOF > pause-pod.yaml
---
apiVersion: scheduling.k8s.io/v1beta1
kind: PriorityClass
metadata:
  name: overprovisioning
value: -1
globalDefault: false
description: "Priority class used by overprovisioning."
---
apiVersion: apps/v1
kind: Deployment
metadata:
  name: overprovisioning
  namespace: kube-system
spec:
  replicas: 1
  selector:
    matchLabels:
      run: overprovisioning
  template:
    metadata:
      labels:
        run: overprovisioning
    spec:
      priorityClassName: overprovisioning
      containers:
      - name: reserve-resources
        image: k8s.gcr.io/pause
        resources:
          requests:
            cpu: 1
            memory: 4Gi
EOF

• apply di cluster

kubectl apply -f pause-pod.yaml

Penutup

Sahabat Blog Learning & Doing demikianlah penjelasan mengenai Understanding and Combining GKE Autoscaling Strategies. Semoga Bermanfaat . Sampai ketemu lagi di postingan berikut nya.