Why you need to know about desktop virtualization network requirements
Desktop virtualization is becoming a reality after years of being touted but rarely used. For networking teams, implementing virtual desktop infrastructure will mean moving swiftly to optimize networks for delivering high-performing desktops and applications, a challenge they've never had to deal with before.
The signs of desktop virtualization uptake are everywhere. While the technology played second fiddle to server virtualization at VMworld 2008, it took center stage at this year's show. That's partially because a number of desktop virtualization pilot projects have moved into actual production, and VMware launched its own virtual desktop infrastructure product, View.
Nemertes Research expects significant enterprise uptake in 2010-2011 -- a marked difference from 2008, when only 22% of companies were deploying desktop virtualization, and 62% of them were still in the evaluation phase.
While desktop support teams find themselves enamored with virtualization because of its cost savings and ease of management, networking teams are stuck between the desktop and the end user -- an unfamiliar and potentially uncomfortable spot.
What is desktop virtualization?
Desktop virtualization is the abstraction of a desktop workload (operating system and applications) from desktop hardware. The agent of extraction is the hypervisor, placed on either the desktop or the server. Virtualization enables multiple desktops to be run on the same machine completely isolated from one another.
Server-centric vs. desktop-centric virtual desktop infrastructure
The dominant mode of desktop virtualization today is server-centric, where virtual machines running desktop workloads sit on servers in a data center. Desktop-centric virtualization runs workloads in virtual machines under hypervisors on desktop/laptop computers.
Desktop-centric virtualization provides offline access to the desktop, since it runs entirely on the machine. Server-centric desktop virtualization requires online access. Some enterprises run a hybrid virtual desktop infrastructure in which users can work online and then switch over to offline when they are ready to leave the office or hop on a plane.
Desktop virtualization benefits
In theory, desktop virtualization is a win-win for IT and the end user. Ideally, to end users, the virtual desktop experience seems identical to a conventional desktop -- an operating system with a suite of applications. Virtual desktops enable users to maintain this same desktop experience even while moving between computers or classes of computers (i.e., desktop, laptop or mobile).
For IT, the upside is that provisioning and management of the OS and applications for many computers facilitates locking down the desktop image. Since 98% of the software doesn't change from one machine to another, system administrators can create a few standard images on which all machines boot. What's more, virtual desktops can always boot off a clean OS with applications and OS patched transparently, simultaneously and comprehensively -- all in the background.
Working to overcome a lack of user personalization
User personalization is a key challenge in desktop virtualization. Starting with a new image is a bit like Groundhog Day -- no record of any previous changes. If the change is a virus, this is a good thing. But if the change is a bunch of new bookmarks for Internet Explorer, it's not so welcome. To address this, all desktop virtualization solutions and a number of third-party providers extract the personal aspects of the desktop and then add them back each time it loads so that it appears just as it was the last time the user logged off.
Desktop virtualization network challenges
But the largest challenge is the ability to guarantee legacy desktop performance in a virtual infrastructure, and vendors differ in their claims of real desktop performance. For example, RingCube claims a 99% performance of a standalone desktop with vDesk, while other vendor solutions are closer to 70% of standalone desktop performance.
The main performance issue is lag time. In the worst-case scenario, there's actually a lag between pressing a key on the keyboard and seeing the letter show up on the screen. This can happen when the network is highly congested. Network congestion can also cause applications to load slowly. For example, users can end up waiting a long time between selecting an application as simple as MS-Word from the start menu and when it actually loads.
It's even more difficult to ensure performance when it comes to real-time applications like voice, video and collaboration. Providing these applications often means adding WAN bandwidth. But these applications and their protocols can also be better supported by implementing traffic prioritization and load-balancing techniques.
In the next piece, learn specific strategies for battling network challenges like bandwidth drain for real-time applications using such methods as traffic prioritization and load balancing.
About the author: Ted Ritter is a senior research analyst with Nemertes Research, where he conducts research, advises clients, and delivers strategic seminars. A Certified Information Systems Security Professional (CISSP), Mr. Ritter leads Nemertes' research on information stewardship, which includes compliance, as well as the management, access, storage and back up of data.
Mr. Ritter has designed, implemented and supported telecom and information security solutions for commercial, federal and international clients. He holds a master's degree in telecommunications management from The George Washington University and a bachelor's degree in neuroscience from Oberlin College.