Vertical Pod Autoscaler (VPA)

1. Introduction to VPA

1.1. What is VPA

VPA is a tool that automatically adjusts vertical resources (CPU and memory) in a container. VPA adjusts Pod’s resource requests based on the actual use of resources in the container, ensuring that the Pod has enough resources during runtime.

1.2. Advantages of VPA

Advantages

Resource Optimization: VPA improves resource utilization by dynamically adjusting container resource requests.
Performance Improvement: VPA enhances application performance and stability by ensuring Pods have enough resources.
Automation: VPA can automatically adjust resources without human intervention.

1.2.1. Limitations of VPA

Since VPA automatically adjusts the resource requests of Pods, Pods may be restarted in the background and may be scheduled to other nodes.
Avoid using VPA and HPA simultaneously.

2. Installation

2.1. Prerequisites

Kubernetes 1.10+

2.2. Deployment

2.2.1. Deploy VPA service certificate


curl -sfL https://docs.surfercloud.com/uk8s/yaml/vpa/gencerts.sh | sh -

2.3. Deploy VPA service

For 1.22 and later versions


kubectl apply -f https://docs.surfercloud.com/uk8s/yaml/vpa/vpa.yaml

For versions below 1.22


kubectl apply -f https://docs.surfercloud.com/uk8s/yaml/vpa/vpa-less-1.22.yaml

2.4. Deploy a VPA object


kubectl apply -f https://docs.surfercloud.com/uk8s/yaml/vpa/deployment.yaml

3. View the VPA object


kubectl describe vpa nginx-vpa

The output of the above command is the value recommended by the VPA for the Deployment.


  Recommendation:
    Container Recommendations:
      Container Name:  nginx
      Lower Bound:
        Cpu:     25m
        Memory:  262144k
      Target:
        Cpu:     25m
        Memory:  262144k
      Uncapped Target:
        Cpu:     25m
        Memory:  262144k
      Upper Bound:
        Cpu:     50032m
        Memory:  107227776429