Field Guide · concept

Also known as: Virtual machine monitor, VMM

A hypervisor is software or firmware that creates and runs virtual machines, dividing one physical computer’s CPU, memory, and I/O among several isolated guest operating systems.1

VM 1 App Guest OS VM 2 App Guest OS VM 3 App Guest OS Hypervisor — creates & isolates the VMs One physical computer · CPU · memory · storage Type 1 sits on the bare hardware; Type 2 runs on a host OS
The hypervisor is the emphasized band in the middle — the engine that turns one machine into many. It carves the host's CPU, memory, and storage into virtual machines and walls each off, so one guest's crash or load can't reach another. A Type 1 hypervisor runs straight on the hardware; a Type 2 runs as an app on a host OS.

Overview

Each virtual machine believes it has its own hardware; the hypervisor mediates access to the real resources and keeps the guests separated. There are two broad kinds. A Type 1 (bare-metal) hypervisor runs directly on the hardware and is what data centers and cloud platforms use for performance and density. A Type 2 (hosted) hypervisor runs as an application on top of an ordinary operating system, which is convenient for desktops and testing. Either way, the hypervisor is the engine that makes virtualization possible.

Where it fits

The hypervisor is what lets one server be sliced into many virtual private servers and underpins infrastructure as a service. It provides stronger isolation than a container, at the cost of running a full guest OS per machine; a bare-metal server deliberately omits it. Running GopherTrunk inside a VM works, though passing a USB SDR through the hypervisor to the guest takes extra configuration.

Sources

  1. Hypervisor — Wikipedia, on Type 1 and Type 2 hypervisors. ↩

See also