I have a deployment with a single replica and a pod disruption budget with minAvailable=1. When I drain the node, the pod does not evict (it is expected and needed) but the node drain process does not schedule the pod on another node. I am looking for a solution so that when I drain the node, the pod should be first scheduled on another node, then should be evicted.
As such affinity is not a requirement, but increasing replica is a requirement, so I can not increase any replica.
I would greatly appreciate any guidance, best practices, or configuration recommendations from the community to achieve this desired behavior in my Kubernetes deployment.
Thank you in advance