So far, we used static PersistentVolumes. We had to create both EBS volumes and Kubernetes PersistentVolumes manually. Only after both became available were we able to deploy Pods that are mounting those volumes through PersistentVolumeClaims. We'll call this process static volume provisioning.
In some cases, static volume provisioning is a necessity. Our infrastructure might not be capable of creating dynamic volumes. That is often the case with on-premise infrastructure with volumes based on NFS. Even then, with a few tools, a change in processes, and right choices for supported volume types, we can often reach the point where volume provisioning is dynamic. Still, that might prove to be a challenge with legacy processes and infrastructure.
Since our cluster is in AWS, we cannot blame legacy infrastructure...
