Server virtualization helps enterprises reduce the total cost of ownership and approach their resources more efficiently. Hypervisors are crucial to server virtualization. In this article, we dissect the two types of hypervisors.
What is a Hypervisor?
A hypervisor refers to software that creates and runs virtual machines (VMs). It enables multiple operating systems to run concurrently and share virtualized hardware resources.
Also known as a virtual machine monitor (VMM), a hypervisor isolates the hypervisor operating system (OS) and resources from the virtual machines. It then allocates resources to each machine and oversees how those resources are scheduled with respect to the physical resource capacity.
The hardware on which a hypervisor runs is referred to as the host, and the virtual machines that use its resources, such as memory and processing, are the guests.
To run virtual machines, hypervisors require operating system-level components like an input/output (I/O) stack, memory and security managers, a network stack, and process schedulers among others.
Types of hypervisors
There are two types of hypervisors: type 1, also known as bare-metal hypervisors, and type 2, also known as hosted hypervisors.
Type 1: Bare-metal hypervisors
A bare-metal hypervisor is a common hypervisor deployment where virtualization software is installed directly on the computing hardware. Since it runs directly on the host machine’s hardware, it does not need an underlying operating system.
These hypervisors are the most efficient, top-performing enterprise computing hypervisors because direct access to hardware ensures they do not contend for virtualization with other software.
Since each guest has its own operating system, the risk of attacks on the vulnerabilities and security gaps that may exist in these operating systems is mitigated. This makes type 1 hypervisors highly secure. In the event a guest virtual machine is attacked, the virtual machine is isolated to ensure other virtual machines on the same hardware are not attacked.
Type 1 hypervisors are suitable for production-level workloads that need constant uptime among more production-ready qualities. They can scale to virtualize workloads spanning hundreds of CPU cores and multiple terabytes of RAM. However, bare-metal hypervisors may incur higher initial costs and require some degree of external support.
As a result, most type 1 hypervisors are suitable for the virtualization needs of large enterprises.
Interestingly, the Kernel-based Virtual Machine hypervisor has a unique model as it is mostly categorized as type 1 but can also be a type 2 hypervisor. It turns the Linux kernel into a bare-metal hypervisor, making it a type 1 hypervisor. As it uses an operating system, the overall system qualifies as a type 2 hypervisor.
Type 2: Hosted hypervisors
A hosted hypervisor is installed atop an existing operating system. It is reliant on the host machine’s operating system to manage calls to resources like memory, CPU, and network. The presence of an operating system between hardware and hypervisor inadvertently introduces latency, even though type 1 and type 2 hypervisors share the same purpose and goals.
Since all hypervisor and virtual machine processes must go through the operating system, bottlenecks are common. Additionally, if the host operating system contains vulnerabilities and security gaps, all the virtual machines running on top of it may be compromised.
These challenges make type 2 hypervisors ideal for use cases where performance and security may as much of a priority as cost. This also makes them ideal test platforms in comparison to the cloud and production virtualized environments.
Type 2 hypervisors usually cost less than their type 1 counterparts. They are unideal for use in data center computing and more effective for end-user/client systems. IT enterprises can use type 2 hypervisors to create virtual desktops.
These hypervisors are suitable for enterprises seeking lower-cost virtualization solutions that are easier to set up and maintain without requiring a dedicated administrator. Latency has a low impact on these environments.
These enterprises also do not need as many virtual machines compared to type 1 use cases.
How to choose a hypervisor
Before choosing a hypervisor, there are a few key considerations to make:
Complexity. Is the deployment and management of the hypervisor simple? Has it been packaged as an independent product with its own console? Does this independence require specialized skills to operate, maintain, or troubleshoot, or is it easy to understand for users with general IT skills?
Performance. A hypervisor should be capable of delivering enough performance for the use cases of your mission-critical applications.
Cost. How do vendors approach hypervisor pricing? Is it open source or would you need to pay for licensing? Is the hypervisor independently priced or is it part of a larger solution?
Ecosystem. A hypervisor that supports popular guest operating systems and leading enterprise technologies and applications belongs to a rich ecosystem.
Why are hypervisors important?
Today, an accumulation of servers that serve independent applications means each server’s resources are underutilized, and this inefficiency has additional implications for the organization’s budget and physical space. Virtual servers, which are enabled by hypervisors, help enterprises use their server resources more efficiently.
Depending on an enterprise’s needs, server virtualization can also provide the following benefits:
- Faster deployment and server provisioning
- Lower hardware costs
- Better scalability and productivity
- Lower energy consumption
- Improved disaster recovery
Enterprises enjoy these benefits of virtualization as hypervisors decouple operating systems and applications from physical hosts. This decoupling makes it easy to migrate virtual machines from one host to another.
Read more: Best Server Virtualization Software