Why do people decide NOT to use VDI today?

Yesterday's conversation on BrianMadden.com was about the real-world non-BS reasons people use VDI today.

Yesterday's conversation on BrianMadden.com was about the real-world non-BS reasons people use VDI today. So I thought a fun follow-up for today would be to have the opposite conversation where we look at the real-world reasons that customers decide NOT to use VDI. Like yesterday, I'll kick off the conversation by sharing the reasons that I've seen customers say "no" to VDI:

Perception that thin client computing doesn’t work

Since VDI is a form of server-based computing which is a form of thin client computing, there are definitely people in the world who believe that thin client computing doesn't work (or that it doesn't work for them). They'll claim that they need multiple monitors or that they have intense graphical requirements of that they need to use their USB accessories.

Unfortunately they don't know that thin client computing has come a long way in the past fifteen years or so, and a lot of things that used to suck about ICA and RDP have been fixed.

Personally I believe that probably 95% of all the world's corporate desktops (not laptops) can be replaced with thin clients. But a lot of people wouldn't agree with that, claiming that thin clients just don't work. (I think it's a variation of NIMBY syndrome.)

Perception that VDI is too complex and/or not ready yet

The big vendors' products are complex. Lots of moving parts. Lots of options. Lots of hype. Not a lot of customers (relatively). There are some smaller players with more turnkey solutions (Pano Logic, for example), but they're more geared toward the SMB space and not everyone's heard of them.

But beyond the product complexity, the whole concept of VDI is complex. We (the collective community) don't quite know if you need application virtualization for VDI. Or disk image sharing. Or offline. Heck, we don't even have a reliable way to predict how many users you can fit on a given piece of hardware.

And if that's not enough, the fact that there are well over 100 vendors in the space doesn't help either. Sure, it's great that our space is so vibrant. But talk about confusion! Of the 100+ vendors, probably 10 are VDI infrastructure vendors, 5 are hardware vendors, and maybe 10 are thin client vendors. That leaves more than 70 companies who could be classified as niche software makers that make some kind of add-on product and tool that will help you with VDI.

So while that's really cool on one hand, it also makes things more complex for customers. Just imagine a new IT admin trying to understand desktop virtualization. If this person attends Synergy, VMworld, or BriForum, he or she would be bombarded with dozens of vendors whose products address the shortcomings of VDI user environment management or printing or monitoring or planning or... And the problem is that each of these products is cool! (Which means it seems like the admin would want them all.) So maybe after all that, the admin just gives up and decides to revisit the technology in the future when the main products have matured.

Perception that it’s too expensive and/or doesn't really save money

Even if a company is not planning to implement VDI just to save money, if the VDI system is too expensive, it's never going to happen.

Acquisition costs are dropping all the time (as we've discussed it before), but there's still not 100% agreement on whether VDI is cheaper to acquire than traditional computers. (Especially once you throw in stuff like VECD.)

And when it comes to operating expenses, it looks like no one really knows whether VDI is actually cheaper to operate than traditional desktops. (Especially since it's so easy to manipulate your cost analysis!)

They use Terminal Server instead :)

Ok, so this one's in here just for me. (See "Madden's Paradox.")

What else?

So that's my short list. What am I forgetting? Have you tried to sell a VDI project (either internally or externally) that didn't happen? If so, what were the show-stoppers?

Join the conversation

17 comments

Send me notifications when other members comment.

Please create a username to comment.

I agree with the above.


Here we came close to not going ahead for various reasons. Institutionalized IT staff that don't like change (the sort that still wear their Novell tee-shirts from the netware 2 days). If we can't even get past this hurdle how are we going to convince the non-it staff, etc. This perception of "whats wrong with what we have now" attitude. Luckily the key for us was to engage with the users and sell them the technology. Demo it to identified users, etc.


We (including me) we're put off by VDI Consultants who despite best efforts to pretend they understand our organisation - don't. They end up annoying you. Getting 2 - 3 phone calls a day from sales losers pretending to know more about us than we do... All this boils into a big wall of negativity. Luckily we closed the door on 99% of solution vendors to allow for breathing space while we did our own workshops/POCs internally.


Of course it didn't help that the big VDI houses were releasing new versions every 5 mins all with features the others didn't have and then they did, etc. It made it very hard to choose a vendor to go with.


Apart from one or two vendors... We had concerns over sustainability, licensing models, etc. Look at the way Citrix changed the model several times last year. That was a major put off.


VDI solutions that require 3rd party products such as Appsense cause major financial concerns. BTW I love Appsense despite not having the need for it at this time. It's certainly the bridge that holds alot of VDI solutions together.


Cancel

Brian,


pretty much agree with your analysis. As a matter of fact I posted an article back in Jan 2008 with the (almost) exact issues:


it20.info/.../90.aspx


Yes.. user-experience (as you noted) has improved since.... all other aspects, unfortunately, got even more complex... (especially complexity and # of vendors).


My 2 cents.


Massimo.


Cancel

In a lot of larger organisations politics comes to the fore here, the separation of the desktop and server teams and the resulting bunfight as to where new responsibilities will lie is simply too much hassle for some.  Note that although this used to happen with Terminal Server, it seems as soon as you put the hypervisor into the equation it gets even worse.


Silly reason I know, but I've still run into this kind of thing more often than I can recall.


Cancel

VDI is too complex for the "desktop team" to grasp and manage. Concepts like ESX, Storage, IOPS are foreign to the bill gates troops out there. They will hold out as long as they can. With Hyper-V they think Hypervisor management is as easy as Windows XP so most Hyper-V VDI projects fail and leave a bad taste in mouths.


The ROI is not there for VDI so the management cannot force the "desktop team" to do VDI.


Cancel

@Jim, completely agree. Politics is a major reason why people don't do virtual desktops or any other new technology.  Desktop virtualization changes the way IT must support the environment.  Instead of each team siloed from the other, you can think of it more like an operating room. Everyone works together to help the patient (user).  Surgeons, nurses, anesthesiologist, etc.  Same for desktop virtualization: servers, networking, app delivery, desktops, desktop delivery,etc.


The second reason is time.  Bigger projects take longer to approve and do which scares many people away.  If a project is going to take a year, many are not sure if they want to do it as it is a major commitment.  If it will take months, not much time is lost if it is a failure.  


Dan (Twitter @djfeller)


Cancel

Customers give us 3 reasons: Money, Time, Complexity.


If VDI can be deployed within a PC replacement budget, if desktop IT can set it up in a few hours, and if it is intuitive for desktop IT to manage, then they are interested.


Kumar (kumar@kaviza.com)


Cancel

What are the issues with complexity?  I feel like there are certain points in time that work perfectly for companies - once they have invested in server virtualization (ESX or Hyper-V) and when the time comes for a PC refresh to migrate to thin clients - what are the complexity problems companies see?


Cancel

1.  Too many parts...


2.  Too many parts vendors...


3.  Too little proven ROI or TCO  (because of #1)


Cancel

Change is evil.. and i might become obsolete..!!


Damn i have to learn new stuff..


Getting out of the comfort zone scares a lot of old "old school" desktop admins and that attitude is a real problem.


Thomas


Cancel

Given that VDI is a nascent market, key question is if VDI WORKS OR IS IT JUST A HYPE?


Our experience at PANO LOGIC indicates that VDI is not only getting adopted, but  users are increasingly scaling up their commitment to VDI.  This is proven by the fact that even as a young player in the market, Pano Logic has 100s of production customers, increasing production deployment of our ZERO endpoint (as opposed to thin/chubby clients) into scores of thousands of units.  Customer satisfaction continues to rise and TCO benefits continue to accrue.


I agree with the observation that thin-client computing doesn't work for VDI. Most of the virtue of VDI architecture is tied to centralizing management of virtual desktop experience.  Thin clients with their ever-increasing chubbiness (sticking multiple processors, OSes, drivers etc. into the endpoint) are reverting VDI back to the ills and challenges of a client-server architecture.  At Pano Logic, our strong conviction is in ZERO clients (no processor, no OS, no drivers, no firmware, no storage at the endpoint) as the right architecture for deploying and benefiting from VDI.  Additionally, bus protocol as opposed to display protocols (that require endpoint processing) are the right delivery medium for task and knowledge worker's VDI experience.


I also agree that VDI solution from most vendors is too complex.  As observed by Brian, Pano Logic provides a solution purpose-built for VDI, and is not an aggregation of various moving parts.  Our customers are consistently deploying our VDI solution (a self-discovered ZERO client endpoint and backend management/brokering software) in less than an hour!  Easy enough.


One of the most important challenge for VDI (versus server virtualization) has been a lack of CAPEX advantage in VDI.  The fact that Pano Logic ZERO endpoint is not burdened with any "tax" to the likes of Microsoft or Intel etc., allows Pano Logic to put the CAPEX proposition for VDI back on the table.  Additionally, 100% centralization of VDI management as well as less than 3 watts power per endpoint delivers a dramatic OPEX proposition.  Net-net, Pano's ZERO endpoint architecture seeks to and will continue to deliver on the TCO proposition for VDI.


===========


Cancel

Cost, complexity and management are the three reasons we hear over and over, and that is because the customer has already been exposed to the VMware or Citrix who skew VDI to their interests.. "desktop virtualization is an extension of server virtualization" in the case of VMware and "VDI needs to be coupled with separate app provisioning infrastructure" in the case of Citrix. Even a Panologic skews the value proposition by bundling the solution with endpoint hardware when it is in the customers best TCO interest to re-purpose existing hardware until it wears out. One of the many value propositions with VDI is breaking the hardware upgrade cycle. And, with our LEAF technology, users can remove their hard drives and turn existing PC's into thin clients that will last another five years.


Virtual Bridges VERDE is one of these few complete VDI solutions that Brian is referring to. In addition, it is not encumbered with VMware or Citrix, or hardware, or even Microsoft VECD, if that is what the customer wants. And, it has no ecosystem dependencies for disconnected use or personalization of the user experience.


Virtual Bridges VERDE also wins on the CapEx side with a stunning $200 provisioning cost (server, storage and VDI) compared to an $800 PC. Throw in VECD for three years and Office and VERDE is equal to the $800 PC, but VERDE includes the management and provision software that negates the $700 per year operating cost required of the traditional PC;


That is how IBM can promote VERDE as part of their GTS SMART Cloud solution for users looking for "ROI's in less that 12 months"


Cancel

I agree with most of what you say, but to be honest there is an overriding reason for a large number of businesses (ours included) why we won't be making a move with VDI (at least yet).


It's almost impossible to make a good business case to migrate away from PC desktops to a 'thin' environment.


We have a virtual infrastructure


We have the central storage strategy


We already use Citrix for delivering some apps.


BUT


PCs are very cheap and these days don't go wrong and require very little management. Even with 2500 of them the support team find it easy to manage.


No matter how smart you are with it, central storage (lot's of it is still required) and that's a cost.


It isn't broken, so we don't need to fix it!


I'm sure we will do it one day (also when the vendors have got their act together and I'm not sure it's there yet) and the business will get the 'same desktop experience from wherever you login' that I presented as the strategic aim. It's just that it costs to make the switch and companies aren't spending right now!


Maybe next year......


Cancel

I agree with Paul Willit's comments. In the past we have never been able to make a good enough business case for the implementation of VDI especially with the majority of our users having laptops. The costs to purchase, implement and support significant server/storage infrastructure has been the main problem.


Maybe the use of VDI in delivering DaaS from the cloud will start to change things.


Instead, we are now looking more at local desktop virtualisation technologies like MokaFive which have a small infrastructure footprint and can run and be supported on non-corporate owned laptops.


Cancel

The main barrier in my experience has always been political and organisational, rather than technical. No one wants to support a VDI infrastructure and an existing desktop estate using completely separate tools, and most existing estates have at least one VDI "deal breaker" - rubbish provisioning system, manual configuration for individual desktops, rubbish application deployment, etc.


The last company I worked for successfully implemented VDI by using the same processes as for physical PCs throughout - including a manual build (post sysprep at least, but manual domain join, locale, etc) .


I have consulted with a number of clients since but the VDI solution only makes sense as part of a complete SOE refresh. I am hopeful that as Windows 7 upgrade cycles start new SOE environments will support Terminal Services, VDI and thick clients with an identical user experience and the same management, deployment and patching tools for the IT department.


Cancel

I think the main problem is that no business of any size is going to ditch an existing PC infrastructure wholesale and implement VDI, so the problem of migrating to and supporting a second environment is always there.


We may get our first opportunity as part of an expansion which will see two new call desks being implemented - and ideal opportunity for a tactical solution, which also servers as a POC.


If we also go ahead with a possible heads office move from one location in London to another then we have the utopia of a 'green field site' and we can really take advantage.


I think that's still a year away and it gives VMware / Citrix etc. time to develop and mature their offerings further!


Cancel

One of the issues that discourage people from using VDI is the performance of media such as Flash videos, display of PDFs or working with graphics-heavy files.  This is especially true for remote users.


One solution to this obstacle is Ericom Blaze, a software-based RDP acceleration and compression product that provides improved performance over WAN and congested LANs. Besides delivering higher frame rates and reducing screen freezes and choppiness, Ericom Blaze accelerates RDP performance by up to 10-25 times, while significantly reducing network bandwidth consumption over low-bandwidth/high latency connections.


Ericom Blaze works with any standard RDP host, including VDI, Terminal Servers and remote physical machines.


You can read more about Blaze and download a free evaluation at:


www.ericom.com/ericom_blaze.asp


Or view a video demo at:


www.ericom.com/blaze_youtube.asp


Adam


Cancel
VDI + GPO = failure
Cancel

-ADS BY GOOGLE

SearchVirtualDesktop

SearchEnterpriseDesktop

SearchServerVirtualization

SearchVMware

Close