Vertical Pod Autoscaler (VPA)

a set of components that automatically adjust the amount of CPU and memory requested by pods running in the Kubernetes Cluster

• Automatically Adjusts CPU and Memory reservations to help right sizing the application
• VPA is not natively available in the Kubernetes Cluster like HPA is. It's necessary to install it manually
• VPA needs the metrics server installed in order to watch the metrics

VPA Components

• VPA Admission Controller
• Whenever a new pod is created it calls this webhook

• It checks whether the Pod (or its Deployment) has a configured VPA

• VPA Recommender

• Connects to the Metric Server to fetch historical and current resource usage for a given vpa-enabled application

• It generates recommendations for scaling up/down the requests and limits for these applications

• VPA Updater

• Updates the resources for an application running out of the recommendation range
• It tells the Admission Hook to change the resource settings before the pod starts
• Runs every 1 minute

Installation

• Check https://github.com/kubernetes/autoscaler

Manifest

apiVersion: autoscaling.k8s.io/v1beta2
kind: VerticalPodAutoscaler
metadata:
  name: my-vpa
spec:
  targetRef:
    apiVersion: apps/v1
    kind: Deployment
    name: my-app
  resourcePolicy:
    containerPolicies:
      - containerName: "*"
        minAllowed:
          cpu: 5m
          memory: 5Mi
        maxAllowed:
          cpu: 1
          memory: 500Mi
        controlledResources: ["cpu", "memory"]