Four reasons why VDI might not be right for you
Virtual desktop infrastructure is supposedly easier to manage and saving enterprises money, but two VDI skeptics wonder whether its limitations are worth the effort.
Don Jones and Greg Shields, of Concentrated Technology, firmly believe that business comes first, not technology. While they feel that virtual desktop infrastructure is undeniably cool, they don't necessarily believe that they offer true value to a lot of the businesses that pursue them. In this series of articles, the skeptics outline their arguments for and against various VDI technologies, helping you focus on what your business actually needs -- not necessarily on what a vendor or two might want to sell.
We'll just say it: Virtual desktop infrastructure, or VDI, is a bad idea.
Full of expensive components, complicated interconnections, layers that haven't fully been realized and a multivendor marketing machine that places sexiness over true business value, VDI represents our industry's red herring du jour. That's a strong statement, but it's intentionally strong to get you thinking.
Why is VDI a bad idea? We'll start with these four reasons.
1. You might not save any time or money
Companies pursue desktop virtualization because they believe it will save them money. For certain users, it might, but too many companies try an "all or nothing" approach and convert all of their desktops to VDI. Here's the problem: If you end up having one virtual desktop per user, then you have achieved nothing other than relocating those users' computers from their desks into the data center. Since those users will still need some kind of access device, you'll still have hardware to troubleshoot and maintain. In addition, many users need to travel with their machines. This will create a situation vastly more complicated and less productive than a simple, cheap laptop computer.
VDI generally only saves money when you can have fewer total virtual desktops than users. You can get this through pooled desktops, which are provisioned on demand. You can also save money where desktop complexity is low, such as in a telephone call center, where each user has a very limited set of applications and functionality, and where computers are not "personalized" very much, if at all.
2. Virtual desktops can cost more than real one
What does a business-class desktop computer go for these days? Ignoring the false costs for so-called management and total cost of ownership, a reasonably-powered one is about $500. Even laptops have fallen below $1,000.
The server hardware needed to virtualize those devices, however, is significantly more expensive, particularly with storage. Unless you can get away with pooling virtual desktops -- that is, having users share a pool of virtual desktops rather than dedicating one to each user -- you're going to spend more money on hardware with VDI than you would if you just bought everyone a desktop or laptop computer.
And those false costs? You can ignore them in the calculation above because they don't go away with the move to VDI. In fact, they stay exactly the same, all on top of that expensive server-class hardware.
VDI promises to offset more-expensive hardware by enabling you to patch and maintain fewer computers. However, the one-desktop-per-user phenomenon doesn't go away without some kind of desktop pooling. Provisioning desktops on the fly is a tricky business, though, and it's only practical when your users' "personal computers" aren't all that personal -- meaning each virtual desktop has minimal customization.
3. VDI converts inexpensive desktop storage to expensive SAN storage
There's a singular problem with the move to VDI: SAN storage. Supremely engineered, ridiculously available, excessively redundant and fantastically expensive, even the lowest-cost SAN storage today can out-cost its equivalent at the desktop by an order of magnitude.
Add up every desktop and laptop in your infrastructure. Then, in the next column, enter in the amount of storage that each device contains. Multiply them, and you're sure to have a giant number. Storage for VDI desktops can occupy terabytes and quickly grow to be one of its biggest costs. Until vendors figure out how to truly de-duplicate primary virtual machine storage, we're stuck with massive storage area networks (SANs) and costly VDI infrastructures.
4. VDI's layers are still too many, too complex
A recent independent technical analysis by our company, Concentrated Technology, found 19 separate components that you must correctly integrate to create VDI. Download a presentation of our findings in the Conference Materials section of our website.
Broken down into no less than 10 distinct layers, the sheer volume of complexity imposed by VDI requires is often too much for the average IT organization to tackle. Enter the expensive consultants. Furthermore, such levels of complexity grow absurdly burdensome to monitor. Is today's performance problem related to a processor shortfall in the virtual host, or is the storage subsystem spindle-bound? Is the user experience suffering because of a transport protocol limitation, or is an application causing the problem? VDI's distinct layers allow for great customizability. But until the VDI "experience" can be better shrink-wrapped (a solution that might not be far off), most environments are better off leaving VDI alone.
Next month: The skeptics explore the business realities that can't be ignored.