Hi, I’m about to start working on enable Horizontal Pod Autoscaler (HPA) and found this article https://sgringwe.com/autoscaling-your-kubernetes-service-12107d7d5f4a suggesting to also enable a Disruption Budget policy. I can be totally wrong but there seems to be some overlap between them, so I was curious to hear what people is doing in terms autoscaling and if that plays nicely with Pod Disruption policies.
I think these two resources (HPA & PDB ) complement each other.
As Cluster Autoscaler and HPA provide a Mechanism to scale up or down nodes and pods based on CPU usage or custom metrics.
On the Other hand, PDB is a policy which defines the minimum number of pods that must remain running when evacuating nodes.
by using both resources In a k8s cluster, defined number of pods will be there despite, nodes will be removed and added.