Virtualization
Virtualisation separates operating systems and applications from physical hardware: a hypervisor divides processors, memory, storage and network into virtual machines that run independently, can be moved while running and can be backed up as a whole. Type 1 hypervisors such as Hyper-V, VMware ESXi or KVM run directly on the hardware; type 2 hypervisors run as an application inside an operating system. Virtualisation is the foundation of consolidation, high availability and every form of cloud.
Virtualization in practice
The practical benefit rests on four points: consolidating many lightly used servers onto few hosts, mobility through live migration of running VMs between nodes, recoverability by backing up whole machines rather than individual files, and isolation between workloads. Technical ceilings are rarely the constraint today: Hyper-V on Windows Server 2025 supports up to 2,048 logical processors, 1,024 running virtual machines and 4 petabytes of memory per host with 5-level paging (256 terabytes with 4-level paging), while a generation 2 VM may use up to 2,048 virtual processors and 240 terabytes of memory, compared with 64 and 1 terabyte for generation 1. A failover cluster holds up to 64 nodes and 8,000 running VMs.
More important than these numbers is the design: generation 2 VMs provide UEFI, Secure Boot and a virtual TPM and are therefore the prerequisite for hardened guests, and consolidation density drives licensing cost, because Windows Server is licensed and Azure Local billed per physical core. A hypervisor alone does not create availability — that needs shared, redundant storage, for instance from a hyperconverged cluster, and a network able to carry live migration and storage traffic at the same time.
Since the licensing changes in the VMware ecosystem many mid-sized companies are evaluating a move to Hyper-V; Alendris assesses such plans in the infrastructure assessment and plans target architecture, sizing and migration sequence.