Assuming that a solution requires a mix of about 70% components running in containers, and 30% (legacy, monoliths, many dependencies, too hard/too expensive to containerize) in VMs, is there a way to manage the VMs via kubevirt when everything is deployed on GKE (well the containers certainly are, while the VMs are on GCP computes, ideally in same availability zone as where the GKE worker nodes are).
The key reason for wanting to use Kubevirt is to be able to treat all resources as K8s resources and orchestrated using K8s.
Is there a way to make this work on GKE, where as a user one is operating within constraints of a managed K8s CaaS ? Any pointers or guidance in this regard would be very helpful.