What Is a Bare Metal Hypervisor?

A bare metal hypervisor is a virtual machine monitor (VMM) that is installed directly onto the hardware of a physical machine. The hypervisor is the layer that sits between the hardware and operating system that makes virtualization possible. In this article, we’ll take a closer look at what a bare metal hypervisor is, how it works, and why you may need one. 

What Is a Hypervisor? 

To gain a deeper understanding of the bare metal hypervisor, we need to take a better look at hypervisors in general. As stated above, hypervisors monitor VMs. They’re also the application by which VMs are created, managed, and operated. 

Hypervisors enable virtualization—which is a technology that allows IT admins to create virtual servers, networks, storage, and other physical hardware as VMs that can all run simultaneously on just one physical server. Hypervisors separate the server’s software, including the operating system (OS) and all of its applications, from the physical hardware on which they reside. They also separate VMs from each other. 

The hypervisor then handles all the necessary transactions by translating requests between the physical machine and VMs. Using multiple VMs, which are detected as guest machines, on a single server helps organizations make the best use of their computing resources, including network bandwidth, storage, and memory. 

There are two primary types of hypervisors: Type 1 and Type 2. A hypervisor’s type depends on how it’s installed. 

Type 1: Hypervisors, or Bare Metal Hypervisors

A Type 1, or bare metal, hypervisor is directly installed onto the computing hardware. Some bare metal hypervisors can even be installed at the firmware level alongside the motherboard’s BIOS. The “bare metal” of this hypervisor type refers to the software being installed directly on the hardware itself. Bare metal hypervisors are the most commonly deployed hypervisors today. 

Type 2: Hypervisors, or Hosted Hypervisors

This type of hypervisor is installed and operated from within the OS of the host machine, which means that the OS controls and manages the physical hardware instead of the hypervisor. The hypervisor is considered “hosted” because it depends on the server’s OS to handle various operations such as managing computing and storage resources. Even though hosted hypervisors operate within the physical server’s existing (or “parent”) OS, VMs can still have additional OSes if needed. This hypervisor typically runs on cloud-based servers that can be shared and accessed by multiple tenants and virtual servers. 

Hosted hypervisors were the first type of hypervisor. Because traditional x86 hardware came with pre-loaded OSes, hypervisors were installed later as software. This introduced latency as transactions and operations had to go through both the OS and the hypervisor. Today, however, the latency isn’t a real problem because hardware acceleration tools are usually available.

How Do Hypervisors Work?

Hypervisors work by providing a layer of abstraction between the underlying hardware resources and the software (operating system and applications) running on top of it. By divvying up the underlying hardware resources such as memory, network, and compute, it’s possible to create multiple virtual machines (i.e., simulated servers) that have access to those resources. The end user installs the operating system and software applications of their choice on top of these virtual machines (VMs) without having to worry about any compatibility issues with the underlying hardware. 

The VMs are known as guests and can operate just as if they were separate computers, each with its own OS, applications, drivers, and so on. The physical server is referred to as the host. 

The hypervisor’s main responsibilities are to create, monitor, and manage the VMs and provision resources as the VMs need them. Hypervisors can also control scheduling of VM operations to ensure the VMs don’t affect each other as they consume resources. When a VM needs CPU, storage, or extra bandwidth to complete a task, it queries the hypervisor and the hypervisor makes it happen. 

Bare Metal Hypervisors vs. Hosted Hypervisors 

The biggest difference between Type 1 and Type 2 hypervisors is the additional layer of software abstraction you have to contend with when using a hosted hypervisor: the operating system of the host machine. Memory, network, and compute resources have to be retrieved through the OS of the host machine, adding latency to requests from VMs. 

When comparing bare metal and hosted hypervisors, it quickly becomes clear that they both have advantages and disadvantages. 

There are several good reasons that bare metal hypervisors are the hypervisor of choice across organizations today: They deliver better stability, scalability, efficiency, and performance than hosted hypervisors, thanks to their direct access to the physical server. They’re also better at resource management. 

Hosted hypervisors can’t match the low latency of bare metal because VM requests have to go through the OS as well as the hypervisor. 

Bare metal hypervisors are typically very secure because they’re separated from the OS, which is vulnerable to attacks. Because hosted hypervisors run within an OS, any OS vulnerabilities could pose a risk to the VMs. 

One area where hosted hypervisors have the advantage is in overall ease of installation, use, and management of the hypervisor and VM environment. Hosted hypervisors are also less expensive than bare metal hypervisors. 

Which Hypervisor Is Best for My Needs?

The quick answer to this question is that it depends on your use case. However, there are some general guidelines followed by most organizations. 

Bare metal hypervisors work best in environments that need high performance, granular management and control, scalability, and strong security. They’re typically used for enterprise applications, server farms, and data centers. 

Hosted hypervisors are ideal for use cases where a bit of latency won’t cause problems. They’re sometimes referred to as client hypervisors because organizations use them in software testing and end-user testing environments. Software developers primarily use hosted hypervisors. 

Pros and Cons of Bare Metal Hypervisors

Bare metal hypervisor pros include: 

• Control: Bare metal hypervisors give you direct access to the hardware, allowing you full control over your system. 
• Best performance: Latency is low due to direct access to hardware, and VMs can operate at the speed of the hardware itself.
• Ultimate security: Separated from the machine’s OS, bare metal hypervisors aren’t affected by the OS’s vulnerabilities.
• Scalability: Bare metal hypervisors enable IT to create sizeable failover clusters and help VMs use physical resources more efficiently. 

Cons include: 

• Higher cost: Bare metal hypervisors are much more expensive than their hosted counterparts, in part because they require hardware.
• Increased complexity: The skills needed to run and manage a bare metal hypervisor are more specialized than for hosted hypervisors and the learning curve can be steep. 

Performant Data Storage for Virtualized Apps

A hypervisor exposes the memory, compute, and network resources of the underlying hardware to its virtual machines, but what about data storage? Most hypervisors do not account for the data storage, only managing the memory needed to perform a task. Your data storage is therefore only as performant as the hardware it’s stored in. If you want truly performant data storage access for your virtual machines, it’s important to invest in bare metal infrastructure that can keep up. 

As a VMware Technology Alliance Partner, Pure Storage has worked closely with VMware to create a truly VM-aware storage solution that integrates seamlessly with VMware vSphere Storage APIs Array Integration (VAAI). 

Discover how Pure Storage can help you uncomplicate data services for VMware hybrid cloud today.