An Overview of the VMware VDI Infrastructure

A Virtual Desktop Infrastructure (VDI) is a networked environment where physical servers host guest virtual machines (VMs) provisioned as desktops for remote users. A hypervisor like the VMware ESXi allows you to create and run guest VMs on a VDI server. Since virtual desktops run from a central location, a VDI is easier to maintain than individual networked PCs. Thus, a VDI is a viable option for beating the costly maintenance cycles of traditional hardware.

What Is a VDI?

A VDI allows organizations to take advantage of the always-on, run-anywhere nature of virtual computing. With a virtual desktop, users are not tied down to the PCs on their desks. Instead, they can use practically any device to access their virtual desktops from anywhere. Thus, a VDI does away with the constraints of traditional office-based computing. With virtual desktops and their data residing in a central location, you avoid potentially destructive system crashes and minimize data theft potential.

With a hypervisor, VDI administrators can create VMs on server hardware, then install guest operating systems such as Windows, macOS and Linux on those VMs. They then provision the virtual desktops with the applications required by their users. Remote users connect to the VMs, boot them up and then work on their virtual desktops.

Since the operating systems and applications in virtual desktops all reside on the VDI server, there are minimal system requirements for the user devices that connect to the VDI. Regular client PCs, thin clients and other devices such as tablet devices and mobile phones can be used to connect to virtual desktops.

VDI server and infrastructure functionality

Hypervisors and VMs play crucial roles in a VDI. When setting up a VDI, pay special attention to the type of hypervisor you plan to use and how your VMs should be organized to meet your specific requirements. You should also implement an interface to facilitate VDI management.

Hypervisors fall into any of two types. Type-1 or bare-metal supervisors, like VMware ESXi and Microsoft Hyper-V, run on top of the server hardware and do not need an underlying operating system to tap into the hardware’s various components, e.g., CPUs and processor cores, memory, networking and storage.

Type-2 or hosted hypervisors, such as Parallels Desktop for Mac and QEMU, run on top of the server operating system, just like any other application. They rely on the server OS to provide them with the resources they need to run.

The ability to access hardware directly means that type-1 hypervisors are faster and more efficient than their type-2 counterparts. The former can marshal all hardware resources at their disposal and apportion them according to need. In the case of the latter, the underlying OS allots resources to the hypervisor instead of the other way around. However, while more powerful, type-1 hypervisors are more complex to implement.

Whatever type of hypervisor you choose, you should see to it that the corresponding hardware can meet the demands you place on the VDI. Nonetheless, hypervisors have a High Availability (HA) feature that allows them to connect to or provision other servers in the event of a catastrophic failure. Implementation of this feature means that when a VDI server goes down, a backup server takes its place. This keeps operational impact to a minimum.

Depending on the setup of your VDI, users can either log on to the same desktop or a different one every time they connect to the VDI. In a persistent VDI, users can personalize or customize their virtual desktops. A persistent VDI closely approximates traditional heavy clients since the state of the virtual desktops are preserved across sessions and users can make changes to their desktops without administrative intervention. The opposite is true for a non-persistent VDI.

If your users work with the same set of application programs daily, a persistent VDI might be best for your organization. Since a persistent VDI provides users with their own virtual desktops, it requires more storage resources. Thus, if consistency is not a requirement, your organization might be better off with a non-persistent VDI.

Other than the above, you would also need to set up a management interface for the VDI. This interface will allow you to create VMs, provision and assign them to users, and back up and add more capacity to your VDI, among other functions.

What Are the Unique Features of VMware ESXi Server?

VMware ESXi is a bare-metal hypervisor. Touted by VMware as the world’s smallest hypervisor, VMware ESXi can directly access your physical servers’ processing, storage and network components, and share these among the VMs in the VDI. Helping with this task is a high-performance cluster file system named Virtual Machine File System.

Otherwise known as vmkernel, the VMware ESXi microkernel orchestrates the smooth operation of the host and its guest VMs. It does this by connecting directly to the CPU and memory and utilizing other modules derived from Linux to connect to the networking and storage components.

Other features unique to VMware ESXi include:

Parallels RAS and VDI Infrastructures

Parallels® Remote Application Server (RAS) is a remote working solution ideal for use within VDI. Parallels RAS provides secure access to your virtual desktops and applications on any device. Its auto-provisioning capabilities mean that you can use it to scale your infrastructure quickly.

Parallels RAS supports VDI providers such as VMware ESXi and VMware Center, Microsoft Hyper-V and Microsoft Hyper-V Failover Cluster. It also supports hypervisors other than VMware ESXi, including Scale Computing HC3 and Nutanix Acropolis.

With Parallels RAS, you can quickly create a secure VDI that can be managed centrally from a single pane of glass. Find out how easy it is to set up your own VDI by downloading the Parallels RAS trial.


References

Wikipedia – Hypervisor

Wikipedia – Hyper-V

VMWare

PhoenixNAP