In my post The Contrarian View on VDI, I tried to outline some of the challenges of pursuing a desktop virtualization project for a company. In the face of all this, what’s the point? Why bother? All this fuss about centralized desktops…. Is it just a technology looking for a problem to solve?
NO!
It’s still good stuff, and the promise of cost savings it there, and I still believe in it, but I have learned that expectations must be set early and repeatedly, and there are several traps to avoid.
Setting Expectations
If you are a company looking to implement centralized or virtualized desktops in the hopes of saving millions of dollars, slow down. I get it, believe me I do…. most companies have healthy yearly budgets for PC replacement, and every new model requires a new image, testing, and rollout. So, over the years most companies have gotten into creating and saving libraries of desktop images.
This is both a blessing and a curse. It’s a curse, of course, because the more images maintained, then the more effort there is to maintain all your desktops. It’s a blessing because the PC form factor is such arch device, with so many interfaces and so much flexibility built in that users are totally empowered to accomplish their tasks in their own idioms.
So, EXPECTATION #1
You will need executive sponsorship.
Realize that the promise of thin clients as a low-cost replacement for the PC requires you buy in to HYPER-STANDARDIZATION. Standardization is generally viewed as a good thing, i have found, as it allows for greatly increased operational efficiency. Most organizations, however, have not yet realized the impact this has on the users. Standardizing on a thin client inevitably means taking significant functionality away from large chunks of the users of the solution, which is never an easy proposition. The users will never agree to it on their own.
EXPECTATION #2
Your cost savings will take 5-6 years.
This one you have to embrace- there’s no way around it. Remember all the software and hardware that is going in to the Datacenter? You are going to increase your costs the first year or two, not decrease them. the savings will come as a result of increased operational efficiency, and reduced travel and effort supporting desktops in the ‘field.’. Is it possible that it will take less time? I suppose so, but I haven’t yet found a company that has proven it out on their books. It would require that you have already quantified ALL the operational costs of all the aspects of supporting your desktops, and i just haven’t seen companies doing that. This can make it difficult to justify…. So,
EXPECTATION #3
Get a Handle on how to Cost Justify the project!
This will mean different things to every company, of course, but in essence i mean you need to understand how and when you will see the savings. They are real, they are there, but it’s not as simple as server virtualization (if I virtualize 50 servers, that is 50 servers for which i don’t need to pay for power and cooling….).
EXPECTATION #4
The experience of the user will change.
I alluded to this in Part 1… the user experience will change. there’s no way around it. Certainly, you will be providing a more robust, secure method of protecting data, not to mention a more reliable and predictable user experience. Many aspects of the user’s method of access may be changing, including the device they are accessing, where the application is executing, how they are printing, where they save files, and so on. Changing any one of these has an impact to their user experience, and changing more than one at a time is the very definition of a “change to the user experience.”. stuff is going to change- get over it.
EXPECTATION #5
It will take longer than you think or want.
Make sure you buffer your project with some slippage. I consider it mandatory that one of the first phases of your project be a desktop assessment. This means you engage a consultant or partner to assist you with figuring out how your desktops are currently being used.n We collectively take for granted all the capabilities of the PC, which means undoubtedly you don’t fully understand all the ways in which they are currently being used by your users.
No one likes being told, “Hey, you don’t know your environment!”. But in this case, you don’t. And believe me, before you roll out a new model for desktop computing, you better understand FULLY exactly what it is you are affecting. A thorough desktop assessment will help you define all your use cases as well as required peripherals. THIS IS ABSOLUTELY NECESSARY before you can select a suitable replacement device, remote protocol, printer support, and whether there are any applications that just won’t work.
EXPECTATION #6
You will need help.
Very few organizations have staff just hanging around waiting for projects to work on. Implementing a VDI solution is a significant undertaking, event for a company with small use case. Furthermore, it is a cross-disciplinary exercise, requiring coordination, design, and new solutions across your storage, networking, server, desktop and security teams. Believe it or not, there are companies out there that have the expertise to do a bang-up job. Do us all a favor, get some help with the design and implementation of the solution – you and your technology partners will be much happier.
In my next segment, “Avoiding Traps,” I will try to outline some common pitfalls that seem to cause companies to take their eyes off the prize.
– Posted using BlogPress…
My Occasional IT Blog... 