Helm chart forces legacy autoscaling manifest

[eahrend]

Hey,

I'm using the latest version of the helm chart, and it's attempting to use the legacy version of autoscaler. Here is my values.yaml:

delegateName: %s
accountId: %s
managerEndpoint: https://app.harness.io/gratis
delegateDockerImage: harness/delegate:24.01.82108
replicas: 2
cpu: 2
memory: 4096
autoscaling:
  enabled: true
  minReplicas: 2
  maxReplicas: 10
upgrader:
  enabled: false
delegateToken: %s
tags: %s

The version of K8S that I'm on is: 1.26.6-gke.1700

[rssnyder]

@eahrend can you make sure you are using the correct and up to date helm repo? I am not able to reproduce this.

install/update repo

➜  delegate-helm-chart git:(main) helm repo add harness-delegate https://app.harness.io/storage/harness-download/delegate-helm-chart/
"harness-delegate" has been added to your repositories
➜  delegate-helm-chart git:(main) helm repo update harness-delegate
Hang tight while we grab the latest from your chart repositories...
...Successfully got an update from the "harness-delegate" chart repository
Update Complete. ⎈Happy Helming!⎈
➜  delegate-helm-chart git:(main) helm search repo harness-delegate
NAME                                    CHART VERSION   APP VERSION DESCRIPTION
harness-delegate/harness-delegate-ng    1.0.18          1.16.0      A Helm chart for deploying harness-delegate

template with my local k8s version v1.28.5+k3s1

➜  delegate-helm-chart git:(main) helm template --namespace harness-delegate-ng --create-namespace \
  harness-delegate/harness-delegate-ng \
  --set delegateName=helm-delegate \
  --set accountId=xxxxxxxxxxxxxxxxxxxxx \
  --set delegateToken=xxxxxxxxxxxxxxxxxx \
  --set managerEndpoint=https://app.harness.io/gratis \
  --set delegateDockerImage=harness/delegate:24.01.82202 \
  --set replicas=1 --set upgrader.enabled=true --set autoscaling.enabled=true --kube-version 'v1.28.5+k3s1' | grep 'autoscaling/v2'
apiVersion: autoscaling/v2

template with your k8s version v1.26.6-gke.1700

➜  delegate-helm-chart git:(main) helm template --namespace harness-delegate-ng --create-namespace \
  harness-delegate/harness-delegate-ng \
  --set delegateName=helm-delegate \
  --set accountId=xxxxxxxxxxxxxxxxxxxxx \
  --set delegateToken=xxxxxxxxxxxxxxxxxx \
  --set managerEndpoint=https://app.harness.io/gratis \
  --set delegateDockerImage=harness/delegate:24.01.82202 \
  --set replicas=1 --set upgrader.enabled=true --set autoscaling.enabled=true --kube-version 'v1.26.6-gke.1700' | grep 'autoscaling/v2'
apiVersion: autoscaling/v2

both template with autoscaling/v2

If the problem persists after a clean repo/update, can you post your results for helm repo list and helm search repo harness-delegate

[eahrend]

FWIW, we're using terraform to deploy these using the helm provider, I'll attempt to do this via helm manually and see what happens.