Problem solve Get help with specific problems with your technologies, process and projects.

Two construction models for endpoint virtualization

There are two ways to construct end-user virtual machines for endpoint virtualization. Persistent and volatile VMs have different management, service and storage requirements.

Endpoint virtualization, the virtualization of end user desktops in central repositories, differs from server virtualization in a variety of ways. One of the main differences is the construction of end user virtual machines (VMs). There are two ways to create desktop VMs, and both methods lend themselves to a different endpoint virtualization model.  

Persistent end-user VMs
When determining your endpoint virtualization model, you first need to consider if you are creating persistent or volatile end user VMs. Persistent end user machines consist of one dedicated VM per user. This VM will include the end user’s required applications, data and user state within itself and requires little in terms of external services to support it. This endpoint virtualization model will, however, require a significant amount of storage space within the central storage fabric because the VM’s disk files will grow with use.

This means the VM’s disk files will require long-term management to help control its consistency and size through the use of defragmentation, compression and virtual disk file maintenance utilities. Persistent VM disk files usually start at 10 GB each and can grow significantly as users add data within the VM.

But persistent VMs do not resolve many of the issues organizations face when running physical desktops. Because end user data is contained within the VM -- as it is within a physical desktop -- the organization must apply various means to protect that data, usually through endpoint virtualization backup tools.

Volatile end-user VMs
Volatile end-user VMs do not have this requirement. That’s because a volatile VM is created on the fly as a user logs in. A volatile VM requires three core components to work properly -- the three core components, in fact, of the bull’s eye endpoint virtualization construction model in part one of this tip:

  • A source desktop VM that contains only the updated operating system and required utilities such as management and anti-malware tools.
  • Virtual applications or applications that have been processed through application virtualization (AppV) tools to capture their running state instead of their installation process. Virtual applications are therefore not installed on the endpoint, but rather streamed to the endpoint -- a process that is initiated by end users when they actually work with an application -- and because of this, can be applied to any VM at user logon.
  • User state virtualization (USV), through either custom tools contained within the endpoint virtualization suite or through standard Microsoft tools such as folder redirection and roaming profiles. USV constantly protects end user data because it stores it outside of the VM. It is also applied when users log on to the VM.

In fact, because the VM includes nothing but the OS and because applications and end user data are applied at logon only, each of the VMs becomes volatile because the VM can be discarded as soon as the user logs off. Therefore, the volatile VM model for endpoint virtualization is much more versatile than the persistent model.

This endpoint virtualization model also requires significantly less storage space because you are no longer storing individual VMs. Instead, you store the source copy of the VM and generate new machines as needed when needed. It’s important to note that both AppV and USV have their own benefits independent of the advantages of endpoint virtualization.

Finally, endpoint virtualization VMs vary in size and complexity based on the type of work the end-user performs. VMware categorizes end users into three different types, each with its own requirements:

  • Task workers focus on a small set of repetitive activities based on a small number of applications. Machine requirements are therefore relatively low.
  • Knowledge workers rely on productivity tools to manage information. They usually require a full productivity suite such as Microsoft Office and they often require access to the Internet. Machine requirements are moderate.
  • Power users work with more complex applications such as development tools or graphical imaging tools. Machine requirements are more strenuous.

Understanding your user requirements is essential when planning for the server resources required to support endpoint virtualization because each VM will require appropriate amounts of RAM, properly sized paging files to reduce disk access and proper virtual disk file sizes.

Consider both models, persistent and volatile end-user VMs, to determine which endpoint virtualization construction is best for your infrastructure.

About the experts
Danielle Ruest and Nelson Ruest are IT experts focused on continuous service availability and infrastructure optimization. They are authors of several books, including Virtualization: A Beginner's Guide and Windows Server 2008, The Complete Reference from McGraw Hill Osborne as well as the MCITP Self-Paced Training Kit (Exam 70-238): Deploying Messaging Solutions with Microsoft Exchange Server 2007 from MS Press. Their latest book is a training kit for Microsoft exam titled 70-652: Configuring Windows Server Virtualization with Hyper-V from MS Press. Contact them at

Dig Deeper on Introduction to virtualization and how-tos

Start the conversation

Send me notifications when other members comment.

Please create a username to comment.