Autoscaling with ServiceInstance

Hello,

I’m trying to figure out how can I set up comput autoscaling using MongoDB Atlas Open Service Broker on Kubernetes.

According to the documentation (here and here), this should work:

apiVersion: servicecatalog.k8s.io/v1beta1
kind: ServiceInstance
metadata:
  name: test-mongo
  namespace: test
spec:
  clusterServiceClassExternalName: mongodb-atlas-aws
  clusterServicePlanExternalName: M10
  parameters:
    cluster:
      autoScaling:
        compute:
          enabled: true
          scaleDownEnabled: true
      providerSettings:
        autoScaling:
          compute:
            maxInstanceSize: M30
            minInstanceSize: M10
        regionName: EU_WEST_1
      providerBackupEnabled: true
      mongoDBMajorVersion: "4.2"

It does set up a cluster without autoscaling.

What am I doing wrong here?

Thanks for the help,

Endre