VMware ESXi and vSphere Cluster Management

What Is a Hypervisor? Types, Architecture, and Virtualization

Learn what a hypervisor is, how virtual machines use hardware resources, how Type 1 and Type 2 hypervisors differ, and how to choose between them.

A hypervisor is software or a virtualization layer that creates, runs, and manages virtual machines (VMs). It is also called a virtual machine monitor (VMM). A single physical computer can use a hypervisor to host one or many virtual machines, each with its own operating system and virtual hardware.

Virtualization converts physical computing resources into software-defined environments. The hypervisor allocates and mediates access to the computer's CPU, memory, storage, networking, and other hardware so that multiple workloads can share one physical system.

For a broader introduction to the subject, see what a hypervisor is.

What Is Virtualization?

Virtualization is the creation of virtual computing environments from physical hardware resources. Instead of assigning an entire physical computer to one operating system, virtualization allows software to present separate virtual computers on the same hardware.

The physical computer is the underlying system. The virtual environments running on it are software-based computers that consume assigned portions of the physical system's resources. The hypervisor controls this relationship and helps prevent one VM from directly taking control of another VM's virtual hardware.

Host, Guest, and Virtual Machine Terms

• Host machine or host system: the physical computer providing resources. In hosted virtualization, “host” can also refer to the computer running the host operating system.
• Host operating system: the existing operating system on which a Type 2 hypervisor is installed as an application.
• Virtual machine: an isolated, software-based computer that uses virtual CPU, memory, storage, and networking.
• Guest: a VM or, more specifically, the operating system running inside that VM.
• Guest operating system: the operating system installed and executed within a VM, such as a Linux distribution or Windows edition.

The terminology changes slightly by architecture. In a Type 2 design, the host operating system is clearly present below the hypervisor application. In a Type 1 design, the hypervisor runs directly on the physical hardware, so there may not be a conventional host operating system in the stack.

What Does a Hypervisor Do?

A hypervisor provides the control layer between physical resources and virtual machines. Its responsibilities commonly include:

• Creating and deleting VMs.
• Assigning virtual CPUs, memory, virtual disks, and virtual network adapters.
• Scheduling VM work on physical processor cores.
• Mapping guest memory to physical memory.
• Presenting standardized virtual hardware to guest operating systems.
• Controlling access to storage, networking, and selected physical devices.
• Starting, stopping, pausing, and monitoring VMs.

For example, a VM configured with two virtual CPUs does not necessarily have two dedicated physical processors. The hypervisor schedules its virtual CPU work on available physical cores along with the work of other VMs and, where applicable, the host operating system.

Isolation and Reliability

Isolation is the separation intended to prevent one VM's workload or failure from directly disrupting other guests or the underlying system. Each VM normally sees its own virtual hardware rather than the complete hardware state of other VMs.

If one guest operating system becomes unresponsive during an experiment, other guests should normally remain available because they are separate execution environments. A web server VM, database VM, and test VM can therefore run on one physical server without being ordinary processes inside one shared operating-system session.

Type 1 Hypervisors: Bare-Metal Virtualization

A Type 1 hypervisor is a bare-metal hypervisor or native hypervisor. It runs directly on physical hardware rather than as an ordinary application on a conventional host operating system.

Physical hardware
        ↓
Type 1 hypervisor
        ↓
VM 1: guest operating system     VM 2: guest operating system     VM 3: guest operating system

In this architecture, the hypervisor is the primary software layer that manages the server's hardware and presents virtual resources to its guests. Type 1 platforms are common in data centers, enterprise servers, private clouds, and environments requiring centralized management of many workloads.

Representative examples include Oracle VM Server, VMware ESX/ESXi, and Microsoft Hyper-V. The exact architecture, management model, driver support, and licensing vary by product.

Example: Consolidating Server Workloads

An organization installs a bare-metal hypervisor on one physical server and creates separate VMs for a web application, a database, and a test environment. The workloads share the server's hardware while retaining separate operating-system environments and administrative boundaries.

Type 2 Hypervisors: Hosted Virtualization

A Type 2 hypervisor is a hosted hypervisor. It is installed as an application on an existing host operating system. The host operating system continues to manage the physical computer, while the hypervisor application requests resources for its VMs.

Physical hardware
        ↓
Host operating system
        ↓
Type 2 hypervisor application
        ↓
VM 1: guest operating system     VM 2: guest operating system

Hosted hypervisors are common for local development, testing, education, demonstrations, and running a second operating system on a desktop or laptop. A developer might install Oracle VirtualBox on a laptop's host operating system and run a Linux guest for software testing without replacing the laptop's primary system.

Representative hosted products include Oracle VirtualBox, Microsoft Virtual PC, VMware Server, and VMware Workstation. Product availability and support differ by version and platform.

Type 1 Compared with Type 2

Bare-metal platforms are generally associated with server and infrastructure deployments, while hosted products are commonly used by individuals on desktop systems. This is a deployment pattern, not a universal performance or security rule. Actual results depend on hardware, configuration, drivers, workload, management features, and product capabilities.

Benefits of Virtualization

• Hardware efficiency: multiple workloads can share one physical computer instead of requiring one computer per workload.
• Workload separation: applications with different operating-system requirements can run in separate VMs.
• Controlled testing: developers and administrators can test software or configuration changes in an isolated environment.
• Portability: a VM's configuration and virtual disks can often be moved between compatible systems.
• Repeatability: standard VM configurations can support consistent labs, development environments, and training exercises.
• Operating-system flexibility: different operating systems can run on the same physical system.
• Simpler development and compatibility work: a developer can test an application against multiple guest environments without maintaining separate physical computers.

Virtualization also introduces operational considerations. VMs compete for shared resources, licensing may apply to the hypervisor and guest operating systems, administration becomes more complex as the number of VMs grows, and every guest and virtualization layer requires appropriate security updates.

Choosing a Hypervisor Type

Start with the deployment goal. Someone testing another operating system locally will often consider a Type 2 hypervisor. An organization operating many always-on server workloads will often evaluate a Type 1 platform. The final choice should be based on requirements rather than type alone.

Virtual Machine Resource Settings

A VM is configured with virtual hardware. Common settings include a number of virtual CPUs, an amount of virtual memory, one or more virtual disks, and one or more virtual network adapters. These are allocations or interfaces managed by the hypervisor, not necessarily dedicated physical devices.

Do not assign fixed resource values without considering the host's capacity and the guest workload. Allocating too much can leave insufficient resources for other VMs or the host; allocating too little can make the guest slow or unstable.

Common Troubleshooting Situations

The VM Is Slow or Unresponsive

• Check whether the host has enough available CPU, memory, and storage capacity.
• Review the VM's resource allocations and adjust them for the guest workload.
• Reduce the number of resource-intensive guests running at the same time.
• Confirm that the host has adequate free disk space.

The Guest Does Not Start After a Configuration Change

• Review virtual CPU, memory, firmware, disk-controller, and other virtual hardware settings.
• Confirm that the intended virtual boot disk is attached and available.
• Check whether the guest installation is damaged.
• Use a known-good installation image or an appropriate recovery process when necessary.

A Hosted Hypervisor Does Not Behave Like a Dedicated Server Platform

This usually results from confusing Type 1 and Type 2 architecture. A hosted hypervisor depends on the host operating system and shares resources with it. Review the software-stack comparison and match the platform to the workload scale, host dependency, and management requirements.

A User Assumes Isolation Is Absolute

A guest crash should normally be contained so that other guests continue operating, but shared resources and security boundaries still matter. Maintain updates, use suitable access controls, monitor resource usage, and avoid overcommitting the host.

Key Points

• A hypervisor, or virtual machine monitor, creates and manages virtual machines.
• Virtualization abstracts physical CPU, memory, storage, and networking into virtual resources.
• The host provides resources; the guest operating system runs inside a virtual machine.
• Type 1 hypervisors run directly on hardware and are commonly used for server and infrastructure workloads.
• Type 2 hypervisors run as applications on host operating systems and are commonly used on desktops and laptops.
• Isolation reduces interference between guests, but it depends on secure configuration, patching, and adequate resources.
• The appropriate hypervisor depends on workload, scale, guest support, hardware, management, networking, storage, licensing, and administrator skills.