How to make HPA roll back the changes when it's deleted?

General Discussions
1

Hi all,

There is a straightforward use case: a new one-replica deploy created as
kubectl create deploy test-metrics --image=nginx --replicas=1
followed by applying the next HPA:

apiVersion: autoscaling/v2
kind: HorizontalPodAutoscaler
metadata:
  name: just-hpa
  namespace: default
spec:
 scaleTargetRef:
  apiVersion: apps/v1
  kind: Deployment
  name: test-metrics
minReplicas: 11
maxReplicas: 20
metrics:
- type: Pods
  pods:
    metric:
      name: requests_per_second
    target:
      type: AverageValue
      averageValue: "1

It led to the immediate growth of the number of pods to 11, as expected.

However, when I deleted the HPA, the number of pods stayed the same because as I found the hpa edited the ReplicaSet of the deployment.

Is there any way to make hpa when it’s deleting to roll back changes?

Thank you!

Cluster information:

Kubernetes version: 1.31
Cloud being used: qemu instances
Installation method: kubeadm
Host OS: Ubuntu 22