The purpose of Microsoft failover clustering is to provide high availability and fault tolerance to the R2 Hyper-V cluster. Alternatively, we can create a VM on the cluster itself, and thereby create a highly available VM. This takes care of the availability of the VM, but failing over would require restarting of the VM on the passive node, which may just—only just— sound feasible in the event of the primary node going down with an acceptable turnaround time. However, in situations involving maintenance of the cluster infrastructure, you cannot afford to take down VMs to load them onto a different node. You migrate them live and running to an other running node (well, until Windows Server 2012) with a negligible—almost unnoticeable—downtime. This is called VM mobility. Windows Server 2012 R2 adds more spice to better an almost perfect recipe, with shared nothing live migration, which does not require the VM to be clustered for mobility...