[Kubernetes] The simple introduction of VPA

VPA stands for Vertical Pod Autoscaling, which frees you from having to think about what values to specify for a container's CPU and memory requests. The autoscaler can recommend values for CPU and memory requests, or it can automatically update values for CPU and memory requests.

Before using VPA, we need to install Metrics Server first as follows:


Prepare the installation of Metrics Server

$ git clone https://github.com/kubernetes-incubator/metrics-server.git
$ vim metrics-server/deploy/1.8+/metrics-server-deployment.yaml
==> Add args in the end
...
...
        args:
        - --kubelet-insecure-tls
        - --kubelet-preferred-address-types=InternalIP

Deploy the yaml file

$ kubectl create -f deploy/1.8+/
$ kubectl -n kube-system get pods | grep metrics
metrics-server-7c9d76cf84-49qzp            1/1     Running   0          4h37m

Verification

$ kubectl top nodes
NAME         CPU(cores)   CPU%   MEMORY(bytes)   MEMORY%
ndc73-0541   2032m        4%     12572Mi         1%

$ kubectl -n kube-system top pods
NAME                                       CPU(cores)   MEMORY(bytes)
calico-node-6kx7t                          39m          345Mi
coredns-858b8fd7d-5b4lb                    4m           21Mi
coredns-858b8fd7d-5h2b6                    9m           21Mi
etcd-ndc73-0541                            29m          95Mi
gpu-upt-n7h6g                              1m           22Mi
kube-apiserver-ndc73-0541                  46m          568Mi
kube-controller-manager-ndc73-0541         51m          78Mi
kube-proxy-55qrg                           11m          53Mi
kube-scheduler-ndc73-0541                  15m          24Mi
metrics-server-7c9d76cf84-49qzp            2m           18Mi
my-scheduler-8d6b544c7-z529w               15m          47Mi
vpa-admission-controller-9c9c8c97d-htnj2   11m          15Mi
vpa-recommender-77c99558b5-b8kzd           11m          16Mi
vpa-updater-59c66946d5-xm49n               13m          16Mi

After finishing the installation of Metrics Server, we may start testing the function of VPA.
Here I follow the official document: Configuring vertical pod autoscaling
https://cloud.google.com/kubernetes-engine/docs/how-to/vertical-pod-autoscaling

1. Getting resource recommendations

my-rec-vpa.yaml

apiVersion: autoscaling.k8s.io/v1beta2
kind: VerticalPodAutoscaler
metadata:
  name: my-rec-vpa
spec:
  targetRef:
    apiVersion: "extensions/v1beta1"
    kind:       Deployment
    name:       my-rec-deployment
  updatePolicy:
    updateMode: "Off"

my-rec-deployment.yaml

apiVersion: extensions/v1beta1
kind: Deployment
metadata:
  name: my-rec-deployment
spec:
  replicas: 2
  template:
    metadata:
      labels:
        app: my-rec-deployment
    spec:
      containers:
      - name: my-rec-container
        image: nginx

Deploy them and see the resource recommendation

$ kubectl create -f  my-rec-vpa.yaml
$ kubectl create -f  my-rec-deployment.yaml
$ kubectl get vpa my-rec-vpa --output yaml
apiVersion: autoscaling.k8s.io/v1beta2
kind: VerticalPodAutoscaler
metadata:
  creationTimestamp: "2019-09-19T07:21:28Z"
  generation: 1
  name: my-rec-vpa
  namespace: default
  resourceVersion: "12429624"
  selfLink: /apis/autoscaling.k8s.io/v1beta2/namespaces/default/verticalpodautoscalers/my-rec-vpa
  uid: 18f042d5-daae-11e9-b99c-ac1f6ba464ec
spec:
  targetRef:
    apiVersion: extensions/v1beta1
    kind: Deployment
    name: my-rec-deployment
  updatePolicy:
    updateMode: "Off"
status:
  conditions:
  - lastTransitionTime: "2019-09-19T07:23:22Z"
    status: "True"
    type: RecommendationProvided
  recommendation:
    containerRecommendations:
    - containerName: my-rec-container
      lowerBound:
        cpu: 25m
        memory: 262144k
      target:
        cpu: 25m
        memory: 262144k
      uncappedTarget:
        cpu: 25m
        memory: 262144k
      upperBound:
        cpu: 7931m
        memory: 8291500k

2. Updating resource requests automatically

my-vpa.yaml

apiVersion: autoscaling.k8s.io/v1beta2
kind: VerticalPodAutoscaler
metadata:
  name: my-vpa
spec:
  targetRef:
    apiVersion: "extensions/v1beta1"
    kind:       Deployment
    name:       my-auto-deployment
  updatePolicy:
    updateMode: "Auto"

my-auto-deployment.yaml

apiVersion: extensions/v1beta1
kind: Deployment
metadata:
  name: my-auto-deployment
spec:
  replicas: 2
  template:
    metadata:
      labels:
        app: my-auto-deployment
    spec:
      containers:
      - name: my-container
        image: k8s.gcr.io/ubuntu-slim:0.1
        resources:
          requests:
            cpu: 100m
            memory: 50Mi
        command: ["/bin/sh"]
        args: ["-c", "while true; do timeout 0.5s yes >/dev/null; sleep 0.5s; done"]

Deploy them

$ kubectl create -f my-vpa.yaml
$ kubectl create -f my-auto-deployment.yaml
$ kubectl get vpa my-vpa --output yaml
apiVersion: autoscaling.k8s.io/v1beta2
kind: VerticalPodAutoscaler
metadata:
  creationTimestamp: "2019-09-19T07:30:14Z"
  generation: 1
  name: my-vpa
  namespace: default
  resourceVersion: "12430493"
  selfLink: /apis/autoscaling.k8s.io/v1beta2/namespaces/default/verticalpodautoscalers/my-vpa
  uid: 52370870-daaf-11e9-b99c-ac1f6ba464ec
spec:
  targetRef:
    apiVersion: extensions/v1beta1
    kind: Deployment
    name: my-auto-deployment
  updatePolicy:
    updateMode: Auto
status:
  conditions:
  - lastTransitionTime: "2019-09-19T07:31:22Z"
    status: "True"
    type: RecommendationProvided
  recommendation:
    containerRecommendations:
    - containerName: my-container
      lowerBound:
        cpu: 586m
        memory: 262144k
      target:
        cpu: 627m
        memory: 262144k
      uncappedTarget:
        cpu: 627m
        memory: 262144k
      upperBound:
        cpu: 777m
        memory: 262144k

The output shows three sets of recommendations for CPU and memory requests: lower bound, target, and upper bound as below with red characters.

The target recommendation says that the container will run optimally if it requests 627 milliCPU and 262144 kilobytes of memory.

Reference:
https://blog.csdn.net/networken/article/details/92830036