Skip to content

Latest commit

 

History

History
44 lines (35 loc) · 1.62 KB

README.md

File metadata and controls

44 lines (35 loc) · 1.62 KB

This template demo the use of Vertical Pod Autoscaler in EKS

Deploy the template

  1. Deploy it using Terraform as you would normally do
  2. Clone the VPA repository which is a submodule 
    git submodule update --init --recursive
  3. Deploy the VPA 
    cd vpa/vertical-pod-autoscaler

./hack/vpa-up.sh
  4. Install Metric Server 
    kubectl apply -f https://github.com/kubernetes-sigs/metrics-server/releases/latest/download/components.yaml

At this stage, we should have something similar

NAME                                        READY   STATUS    RESTARTS   AGE
[...]
metrics-server-8459fc497-kfj8w              1/1     Running   0          83m
vpa-admission-controller-68c748777d-ppspd   1/1     Running   0          7s
vpa-recommender-6fc8c67d85-gljpl            1/1     Running   0          8s
vpa-updater-786b96955c-bgp9d                1/1     Running   0          8s

Test the VPA

The step above will deploy a Deployment, and a Vertical Pod Autoscaler that selected the Deployment as its target. The Vertical Pod Autoscaler will adjust the resource requests of the Deployment based on the observed utilization of the pods.

In the example, the Pods in the Deployment are configured to with 100m CPU requests and 50Mi Memory requests, while the Pod actually use ~ 500m CPU.

After a while (in a few minutes), VPA will adjust the resource requests and launch new pods.

Run the following command to see the VPA status

kubectl get vpa/hamster-vpa -n <namespace>

Reference