i am running a job in my AKS cluster with two node pools : user and system node pool
i also want to scale down the user node pool to zero(0) after my job is completed.
what i did:
1) i enabled the cluster-autoscaler and the node is scaling up.
1) the node is not scaling down to zero because the job is still active (seeking azure queue storage for next triggers).
2) when i delete the completed job manually then only the node is scaling down to zero. but my requirement is not to do manual task here i want that the job will schedule to the system node once completed and the again come back to user node when it receives triggers.
please suggest me some way through which i can achieve that.