Kubelets are agents of Kubernetes which offload workloads, providing abstraction for compute and memory by managing the lifecycle of containers, monitoring resource usage and ensuring that the necessary compute and memory resources are allocated appropriately. Now that we understand this, we can move into understanding Virtual Kubelets further and how to use them efficiently with the help of a Virtual Kubelet Provider.
No, there is no such thing as a supernatural kubelet. We’re still talking about Virtual Kubelets here. If virtual kubelets are ghosts of a node and have no real resources from the cluster, the cluster. They must be deriving their powers from somewhere to be useful in the re, they are drawing powers from a remote/distant source of energy and running the cluster workloads assigned to them. Kubelets are strict interfaces to resource utilization through container runtimes. Virtual Kubelets are just the interfaces. One can write a provider using virtual kubelet to allow users to interface resource utilization of a disparate system into Kubernetes or use these interfaces for simulation.
We are going to try using such a Virtual Kubelet which can delegate to a remote Virtual Machine where our workloads will run. For this we are going to setup Interlink.