What is a virtual machine?

Users can choose from two different types of virtual machines—process VMs and system VMs:  

A process virtual machine allows a single process to run as an application on a host machine, providing a platform-independent programming environment by masking the information of the underlying hardware or operating system. An example of a process VM is the Java Virtual Machine, which enables any operating system to run Java applications as if they were native to that system.   

A system virtual machine is fully virtualized to substitute for a physical machine. A system platform supports the sharing of a host computer’s physical resources between multiple virtual machines, each running its own copy of the operating system. This virtualization process relies on a hypervisor, which can run on bare hardware, such as VMware ESXi, or on top of an operating system.  

All the components of a traditional data center or IT infrastructure can be virtualized today, with various specific types of virtualization:   

• Hardware virtualization: When virtualizing hardware, virtual versions of computers and operating systems (VMs) are created and consolidated into a single, primary, physical server. A hypervisor communicates directly with a physical server’s disk space and CPU to manage the VMs. Hardware virtualization, which is also known as server virtualization, allows hardware resources to be utilized more efficiently and for one machine to simultaneously run different operating systems. 
• Software virtualization: Software virtualization creates a computer system complete with hardware that allows one or more guest operating systems to run on a physical host machine. For example, Android OS can run on a host machine that is natively using a Microsoft Windows OS, utilizing the same hardware as the host machine does. Additionally, applications can be virtualized and delivered from a server to an end user’s device, such as a laptop or smartphone. This allows employees to access centrally hosted applications when working remotely. 
• Storage virtualization: Storage can be virtualized by consolidating multiple physical storage devices to appear as a single storage device. Benefits include increased performance and speed, load balancing and reduced costs. Storage virtualization also helps with disaster recovery planning, as virtual storage data can be duplicated and quickly transferred to another location, reducing downtime.  
• Network virtualization: Multiple sub-networks can be created on the same physical network by combining equipment into a single, software-based virtual network resource. Network virtualization also divides available bandwidth into multiple, independent channels, each of which can be assigned to servers and devices in real time. Advantages include increased reliability, network speed, security and better monitoring of data usage. Network virtualization can be a good choice for companies with a high volume of users who need access at all times. 
• Desktop virtualization: This common type of virtualization separates the desktop environment from the physical device and stores a desktop on a remote server, allowing users to access their desktops from anywhere on any device. In addition to easy accessibility, benefits of virtual desktops include better data security, cost savings on software licenses and updates, and ease of management. 
  

Like virtual machines, container technology such as Kubernetes is similar in the sense of running isolated applications on a single platform. While virtual machines virtualize the hardware layer to create a “computer,” containers package up just a single app along with its dependencies. Virtual machines are often managed by a hypervisor, whereas container systems provide shared operating system services from the underlying host and isolate the applications using virtual-memory hardware.  

A key benefit of containers is that they have less overhead compared to virtual machines. Containers include only the binaries, libraries and other required dependencies, and the application. Containers that are on the same host share the same operating system kernel, making containers much smaller than virtual machines. As a result, containers boot faster, maximize server resources, and make delivering applications easier. Containers have become popluar for use cases such as web applications, DevOps testing, microservices and maximizing the number of apps that can be deployed per server.  

Virtual machines are larger and slower to boot than containers. They are logically isolated from one another, with their own operating system kernel, and offer the benefits of a completely separate operating system. Virtual machines are best for running multiple applications together, monolithic applications, isolation between apps, and for legacy apps running on older operating systems. Containers and virtual machines may also be used together. 

Virtual machines can be simple to set up, and there are many guides online that walk users through the process. VMware offers one such useful virtual machine set-up guide.