What is a Hypervisor and How it Works? We explain the different Hypervisor types. At the epicentre of virtualization sits a well known but seldom discussed technology called the hypervisor. It is a piece of application, firmware, or hardware that makes virtualization possible.
The hypervisor is a layer of software that enables single hardware to host several remote virtual devices. Additionally, it helps with managing virtual machines. Before we learn more about hypervisors, how they work, and their types, let us start with a technology that is closely related to hypervisors virtualization.
What is Virtualization?
The virtual form of a resource such as a desktop, a server, storage space, files, a network, or an operating system is called virtualization. You can transform traditional computing with virtualization since the technology transforms the resources into scalable resources per the organization or user’s needs.
Virtualization is not the latest concept and has been around for years. It is the concept of changing the thought process from physical to logical. Virtualization is all about creating more logical IT resources, called “virtual systems,” within a single physical system. It can be segregated into server virtualization, Operating System (OS) virtualization, and hardware virtualization.
So how does the hypervisor fit into virtualization? Let us understand the hypervisor.
What is Hypervisor?
The hypervisor is a segment of an application, hardware, or firmware that creates virtualization. The technology abstracts and isolates different operating systems and applications from the primary computer hardware. Hence, the functions of the hypervisor enable virtualization. It creates virtual layers to separate physical components like RAM from virtual machines.
A computer that runs and supports the hypervisor is known as the ‘host’ and the virtual machines created are known as the ‘guests.’ In other words, the actual hardware, also known as the host, can independently run and operate one or more virtual machines also known as guest machines.
The hypervisor helps manage the independent guest machines by distributing hardware resources like CPU usage, network bandwidth, memory allotment, and more. This is done by creating pools of hardware resources and allocating them to virtual machines. It can also start and stop virtual machines when requested by the user.
Hypervisor ensures that virtual machines stay isolated from one another. If there is a problem in one virtual machine, the others are unaffected. Also, hypervisor handles the communication of virtual devices over virtual networks. Thus, enabling virtual machines to connect with others.
Since hypervisors are tasked with managing and allocating resources to guest machines, they also allow multiple operating systems to run simultaneously while utilizing the same virtual resource. Some hypervisor vendor examples are Microsoft Hyper, Xen, VMware, and Kernel based Virtual Machine (KVM).
How Does a Hypervisor Work?
With a hypervisor, you create and manage virtual machines by producing a layer between hardware and software. Hypervisor makes virtualization possible by translating requests from physical to virtual resources (RAM, CPU, and Storage) and vice versa.
To understand how hypervisor works, it is essential to know the different types of hypervisors. How do these types work, and what are the differences.
Hypervisor Types and Examples
There are 2 types of hypervisors. They are Type 1 Hypervisor, also referred to as bare metal or native hypervisor, and Type 2 Hypervisor or hosted hypervisor.
Hypervisor Type 1 - Bare Metal or Native Hypervisor
Type 1 hypervisor, or native hypervisor, runs directly on the host machine’s hardware without any intervention from an operating system. The hypervisor has direct access to the hardware without participating with the drivers and the operating system.
The native or bare metal hypervisor is considered to be the most efficient and best performing hypervisor for enterprise computing. This hypervisor is more scalable because of its ability to assign resources directly. There are more advantages to Type 1 hypervisor. They include:
Optimization of physical resources: Enterprises often exhaust budgets quickly by purchasing separate servers for different applications. It is an endeavor that takes up data center space and is time consuming. When a Type 1 hypervisor allows IT to use server hardware, it reduces data center expenses, real estate costs, and energy consumption.
Greater resource allocation: In most Type 1 hypervisors, admins are given the opportunity to manually set resource allocation based on the priority of the application. Several native hypervisors automate resource allocation as per requirement, allowing resource management to be customized and dynamic options.
Some examples of Type 1 hypervisors are as follows:
VMware ESX and ESXi – These hypervisors offer scalability and advanced features but require licensing. Hence, the cost price is on the higher side. There are some low cost bundles that VMware offers and makes hypervisor technology affordable for small infrastructures. VMware is the market leader in Type 1 hypervisors. The vSphere/ESXi products are available in free and five commercial editions.
Microsoft Hyper V – Unlike VMware, Microsoft Hyper V does not offer many advanced features. However, it is one of the top 3 Type 1 hypervisors in the market. The product was first released with Windows Server. Still, now, it has been highly enhanced with the latest version of Windows Server Hyper V Nested Virtualization. Hyper V is available in free edition and four commercial editions – Foundations (OEM only), Essentials, Standard, and Datacenter Hyper V. The free edition comes with no virtualization and no GUI.
The other Type 1 hypervisors are Oracle VM and Citrix XenServer. Both lack the advanced features found in native hypervisors.
Hypervisor Type 2 - Hosted Hypervisor
Type 2 hypervisor is built on top of the operating system. Type 2 hypervisor is in direct contrast to Type 1 hypervisor because it relies on the operating system of the host machine. It is also called hosted hypervisor. Here, the hypervisor runs as an application with the operating system that runs directly on the host device.
The hosted hypervisor or Type 2 hypervisor supports multiple guest machines but is restricted from directly accessing the host hardware and resources. The already existing operating system manages the calls to the CPU for network, memory, storage, and resources. This can, however, create a certain amount of latency. But, such scenarios are restricted to more complex and high performance cases.
The Type 2 hypervisor comes with its set of benefits:
- They are much easier to manage and set up since you already have an operating system to work with.
- A dedicated admin is not required.
- Type 2 hypervisor is compatible with a gamut of hardware.
Some examples of Type 2 hypervisors are as follows:
VMware Workstation/Fusion/Player – VMware Player is a free virtualization hypervisor intended to run only on virtual machines. It does not allow creating VMs. VMware Workstation is a powerful hypervisor with advanced features like VM snapshot support and record and replay.
Kernel based Virtual Machine (KVM) – KVM – It has both the qualities of a Native hypervisor and a Hosted hypervisor. It can turn the Linux kernel into a hypervisor so that the virtual machines have direct access to the physical hardware. You can add the open source KVM to most Linux operating systems, including Debian, Ubuntu, SUSE, and Red Hat Enterprise Linux. You can also add it to Windows and Solaris.