Cluster autoscaler priority expander to use spotpool

[veldpworld]

Hi All,

we would like to use azure spot instance in nonprod environment. we use preferrednodeschedling affinitiy to schedule on spot and it will fallback to ondeman pool in case spot not available. but autoscaler not treat the preferred affinity to scale the node. hence i tried using the priority expander mode autoscaler with highes priority is spot pool name and lowest priority is ondemand. this is not properly increase the spot for some reason , CA adds more node in ondemand.

az aks update \
  --resource-group "rg-example-dev" \
  --name "aks-example-dev" \
  --cluster-autoscaler-profile "scan-interval=20s,scale-down-delay-after-add=5m,scale-down-utilization-threshold=0.6,scale-down-delay-after-failure=1m,scale-down-unneeded-time=5m,scale-down-unready-time=5m,max-graceful-termination-sec=30,skip-nodes-with-local-storage=false,skip-nodes-with-system-pods=true,balance-similar-node-groups=false,max-empty-bulk-delete=10,max-node-provision-time=15m,new-pod-scale-up-delay=10s,expander=priority"


apiVersion: v1
kind: ConfigMap
metadata:
  name: cluster-autoscaler-priority-expander
  namespace: kube-system
data:
  priorities: |
    100: spotpool         
    50:  ondemandpool     

Please let me the right approach or configuration to acheive this. thanks