HorizontalPodAutoscaler Walkthrough

This example is based on horizontal pod autoscale example. https://kubernetes.io/docs/tasks/run-application/horizontal-pod-autoscale-walkthrough/

Compared to the original example, this one is based on golang and it is using an unprivileged port 1025 to be compliant with pod security policies.

Install The Example Application

With the following command, the sample application will be installed.

❯ kubectl apply -f webserver.yaml

To start the load test use:

❯ kubectl run -it --rm load-generator --image=busybox /bin/sh --generator=run-pod/v1
while true; do wget -q -O- http://webserver-service; done

The scheduler will scale up, until the cpu usage is below the defined 20%.

❯ kubectl get hpa
NAME         REFERENCE               TARGETS   MINPODS   MAXPODS   REPLICAS   AGE
webserver    Deployment/php-apache   19%/20%   1         100       7          17m