Making your case for hosted VDI
Over the past few years, virtual desktop infrastructure (VDI) has received some serious press, as well as banter among those in the technology community. The topics range from the different vendors and their technology bake-offs, use cases and best practices, and the debate around total cost of ownership. In addition to the obvious benefits of centralization and simplified administration, the deployment model and technologies that power desktop virtualization have matured significantly. For many companies, the total cost of ownership benefits, if measured properly, are undeniable when coupled with all the other efficiencies that desktop virtualization enables.The path to VDI The industry has some general terminology for buying desktops and centralizing the hosting of business applications as a subscription-based service that you should be aware of before you start evaluating VDI solutions. The two most widely used terms are hosted VDI and Desktop-as-a-Service (DaaS). While there are technical and performance differences between the two (for a closer look, check out the VDI decision matrix), most major decision points are the same. Where these models differ is primarily around resources (shared or dedicated), control, management, and flexibility. While these have a very real impact on the end-user community and customer satisfaction, the business argument for both is very similar.When talking with potential customers, one of the first questions I’m asked is: Why virtualize my desktops at all? This is a rhetorical question. After all, the customer wouldn’t have come to me in the first place if they didn’t have a preconceived expectation of VDI providing greater IT efficiency and productivity gains, reduced administration, or lower total cost of ownership.I use this question as an opportunity to right-size expectations in the front end of an evaluation process, because if the organization’s idea is simply to save money, then it really is approaching the process incorrectly. Sure, if analyzed properly, there will certainly be cost advantages to deploying VDI in a hosted model versus building your own VDI architecture.But these will come from a variety of expense areas, and you need to be sure that you can cover them all. So the real question you should be asking yourself is: Can I get more value (feature, functions, resources, etc.) for the exact same dollar by deploying a hosted VDI model over attempting to do this on my own? In short, you’re looking to get more for your money; the primary driver should not be savings.To determine the answer, I help customers analyze their existing traditional environment. The traditional environment means keeping the physical machines that you paid a lot of money for, a single operating system that is tied to that physical machine, and your applications that are also tied to that single operating system and machine. End-user experience and performance is tied to what? You guessed it … that same physical machine.Also included in the traditional environment is the other gear and applications that sit in a closet somewhere else connected together by wires, switches, ports, and plugs. At this point, you’re probably thinking, “Wow, that’s a lot of moving parts — this must be a management nightmare!” If that’s the case, then virtualizing the IT and desktop environment is a good option for you, because it will centralize and virtualize many of the boxes, apps, ports, and plugs.Once you’ve decided to virtualize, it’s time to decide whether to build your own VDI architecture or align with a service provider that has already built one, completed the heavy lifting, and has a team of competent and experienced professionals on staff.Building your own VDI architecture Build-your-own VDI will involve many aspects of your traditional environment. The difference is that you will be centralizing most of the desktop software that your users interact with along with the applications in the data center. This means that you are now sitting resident (local) with all your line-of-business applications (CRM, Accounting, ERP, Document Management, etc.) on the data center high-performance computing gear.You would then access these centralized desktops through a variety of remote display technologies to paint the “desktop” to your physical devices (laptops, netbooks, tablets, and mobile devices). Think of it as a window into the data center that you have access to from everywhere. A PC, laptop, or tablet is simply a place to access your window.When making any decision, the best way to determine what’s right for you is to use the tried-and-true list of pros and cons.Pros of build-your-own VDI:
Cons of build-your-own VDI:
- Centralized management.
- Simplified application management and upgrades. Install once and deploy many. For example, if you need to migrate from Windows XP to Windows 7, it would be much easier to install this on one virtual desktop, tune that single desktop image, then clone (make an exact copy) for your other 199 users than have to physically touch every single PC or laptop to perform the upgrade.
- Simpler and tighter security.
- Mobility. The ability to enable employees to access a single desktop from many different places.
- Lower total cost of ownership.
- Easy changes to the desktop environment. Let’s say, for example, that you need to install a new application on 75 desktops that is going to require a 2GB memory upgrade. Collecting all 75 desktops and performing 75 physical memory upgrades can be time consuming. In a Virtual Desktop Infrastructure environment, your memory upgrades are reduced to a few mouse clicks at a central location one time and pushing this out to 75 users.
You’ll notice the cons list is almost equal to the pros list. This is the conundrum that IT and financial decision makers have faced for years: How do we acquire these enormous benefits of desktop virtualization without the trade-off of more upfront cash and a heavier IT burden? Enter hosted VDI.Hosted VDI Put simply, hosted VDI is all of the pros without most of the cons. Rather than undertake the task of building your own VDI, you rent the technology and expertise from someone who knows what they’re doing. You’ll save yourself the heavy lifting and have access to a bunch of propeller heads already on staff, while still gaining all the IT efficiencies, maintaining a large portion of control, and leveraging the flexibility of an enterprise-grade platform built for service-provider scale.You’ll also pick up some additional benefits in this model that make a ton of sense financially and need to be quantified in your total cost of ownership evaluation:
- Staffing extras. If you don’t have the staff with the knowledge to build, manage, and administer this new infrastructure, then you’ll pay a hefty penny either on acquiring the skills from the outside or paying a third party.
- More cost to the data center. Deploying this type of architecture adds to the cost of the data center — computing (processing, memory, and storage), licensing, management, and definitely complexity.
- Dependency. Once you centralize computing, you become dependent on high-availability infrastructure, redundant equipment, network, and the staffing requirements to keep things running.
After walking through all the options, I want to make it clear that I’m not saying that a build-your-own data center is a bad idea. I’m simply pointing out that you have an alternative. And if you are considering a virtual desktop or application deployment, you might find that for the same dollar amount you can cloud-source and get a lot more value, features, and functions than trying to plow this road on your own.For a more detailed look at all the differences between the VDI options on the market, refer to the Evolve IP Hosted VDI decision matrix at www.evolveip.net/vdi. Categories: Virtualization
- High availability. Let’s face it, if you are a company with 30 to 1,000 employees, it’s highly unlikely you will ever employ the same level of diversity across all data center elements and networks that a service provider bakes into the offer. It’s just not realistic.
- High redundancy. You pick up access to platform redundancy that is built for guaranteed uptime and availability. To be direct, it’s built from the ground up to withstand failures, weather events, and deliver near-zero Recovery Time Objectives (RTOs). Most folks I talk to in the mid-market build data centers to be back online in 24 hours or more in a disaster recovery scenario. For a service provider, there is no such thing as “recovery time.” We build to withstand failures from the ground up and expect to run uninterrupted during catastrophes.
- Better use of your existing IT resources. Free up IT’s time to be more efficient and productive for the operations of the business, rather than dealing with the plumbing. You also aren’t bringing on additional Full Time Equivalents (FTEs) to manage this new environment.
- No capital. You can pay as you go, scale up, or shrink down. The bill changes with you.