dcsimg

The Short Life Expectancy of the Virtualized Desktop

RE: Current technology for desktop virtualization -- you're doing it wrong.

Current virtualized desktop solutions have the life expectancy of a fruit fly. Even in computer terms that is very, very short. So short, in fact, that you may want to bypass it completely. So, Mr. Virtualization Guy, what’s your brilliant long-term solution for the Desktop dilemma? Web-based, vendor neutral desktop interfaces—sometimes known as webtops or net tops. There, I said it and somehow I know the hairs on the back of your neck are standing at attention and you’re getting ready to flame me. Allow me to pre-defend myself.

Current technology is slow, heavy, and only maybe a little cheaper than standard local desktop operating systems. I think that these heavy desktop virtual machine solutions are generally no better than a standard desktop operating system running on that humming hulk under your desk. How can I say that when I seem to be such a proponent of VDI (Virtualized Desktop Infrastructure)?

I’m a proponent of VDI but common sense tells me that it is not a long-term solution to the expense of desktop support, software licensing costs, patching, backups and so on. VDI is a transitional solution and you’re not getting rid of those issues; you’re just moving them to the data center and from desktop support folks to server administrators. I have been in the IT business long enough to know that people (companies) resist change and when they do change, it is at a very slow pace. Most changes take place in baby steps. VDI is one of those baby steps. It is the transition from traditional desktops to go anywhere web-based desktops.

We’ve spent years web-enabling our applications, creating virtualized server infrastructures, and developing Web 2.0 technology—to what end? Are you going to tell me that we’re going to continue to use 2GB, 3GB (or larger) heavy desktop operating systems and continually boost our LAN and WAN network bandwidth just so we can run those heavy desktops to use our web applications? Really?

A web-based desktop is a brilliant, scalable, inexpensive solution to a very difficult problem: What to do with the desktop. Put your desktop on the web, let your users choose which interface look and feel they want to use (no more religious wars between Mac, Windows, or Linux zealots), use online applications like Google Docs, Zoho, or even stream your own purchased ones for the users that want them.

Most users don’t care which operating system they use and many of them don’t know what they’re using anyway. What they really want is applications and more than 90% of all users use; a word processor, a spreadsheet, a web browser, and email. Why not remove the OS question completely by just using a browser for everything—even the Desktop itself? How much bandwidth would you need if all your users only used a web browser—for everything?

To see what I’m talking about, I’ve provided some links to hosted desktop sites below so that you can check them out for yourself. There are others but these are the best of breed in this space. Try them out and let me know what you think.

StoneWare WebOS
Global Hosted Operating System
iCloud

Comments on "The Short Life Expectancy of the Virtualized Desktop"

esj

not everyone have working eyes or hands. browsers can not run accessibility software such as NaturallySpeaking or Jaws. Accessibility software needs raw machine access because they are near realtime applications as are other audio apps. the end result of a web desktop is exclusion of disabled people from jobs and online services.

curtiscu

I think your article demonstrates a lack of knowledge of VDI technology. The paradigm change is significant, the reduction in effort required for desktop administration is quite substantial.

For example, if you’re delivering a “standard” (whatever that means for your business) web enabled desktop to a thin client, and the desktop becomes corrupt, it takes minutes to roll out a fresh desktop from a preconfigured image in the data center. The alternative of waiting for a support rep to physically arrive at the desktop (worst case scenario) to repair things is orders of magnitude slower.

How is this shift *not* a benefit?

Full disclosure: I’ve worked with software and hardware based thin client technologies for a total of 10 years.

neiliss

at the present, iCloud only supports IE on Windows.

mahaavedha

Are you telling me that i am going to trust all my data to some cloud computing vendor??? you must be kiddng !!

smbulkley

This is the first IT article I’ve read that tried to make central points of administration sound like a BAD thing.

khess

I fully understand VDI technology–I used to be a Desktop Support Tech. I also know that current VDI technology is not going to really solve that much for the user but I do understand–for example, in my article from Sys Admin Nov. 2002: Linux As A Windows Terminal Server Client, my solution then (6 years ago) was a type of desktop virtualization and Terminal Services is still a good technology but you still can’t get hundreds or thousands of users on a Terminal Server setup.

Why do you want to manage 5,000 desktop computers and 1,000 servers, 300 network appliances, etc.? The answer is that you don’t have to and you don’t have to allow some cloud vendor manage it for you either. If you checked out StoneWare’s WebOS, you’d realize that you can do it.

The bottom line is that the traditional desktop is going to die. Heavy client computing has terminal bloat disease and a web-based solution is the only solution and it saves tons of bandwidth. Think about it.

bryanjrichard

I’ve been seeing this argument pop up in response to services in the cloud and I really don’t think it’s going to stand the test of time. It seems like it’s just too snap of a judgment on a relatively new technology. Like saying MySQL “is [only] being used for something that’s tiny — to keep track of your recipes.

A bit shortsighted, that.

I’m not terribly confident in my ability to maintain the integrity and security of the data on my laptop. And my son’s. And my daughter’s. And my wife’s. There’s too much data on too many devices and too little time.

I need all of this nonsense centralized and managed by someone else. I need the cloud.

alizard

People will do what works for them. Back in the early 60s, there was an American pop-rock sound and everybody listened to American Bandstand to find out what it was. Now, if you see 5 different high school kids with MP3 players, it’s likely as not that they’ll be listening to 5 radically different genres of music.

The idea of “one big computing paradigm for everybody” will be equally obsolete Real Soon Now.

For me, my only real dissatisfaction with Virtualbox and the 5 different OSs I run on it is that I can’t buy a copy of OSX that will legally run on it. My goal is to be able to run whatever app I need that best fits what I need to do with the data in the datapool common to all of the above. I also work on proprietary projects, meaning I’m unlikely to trust my data to any cloud I do not own personally.

The home user who wants to websurf / e-mail / IM / light word processing will probably be happy with what you describe, though there’s evidence against it based on the dismal sales of web appliances. Maybe higher bandwidth will help. Maybe OSs designed with webapps in mind will help.

There are lots of places in the business and enterprise environment where people will need both heavy metal desktops and virtual desktops on the Web.

As for Windows Terminal Services … I expect to be getting a mininotebook shortly and using rdp to connect to my home Linux box with it… to run industrial strength apps that would hammer my mininotebook flat. Why rdp? Built-in support for encryption vnc lacks. But if I’m websurfing on the road, what do I need to run my desktop for?

I use the tools that suit my needs, just like everyone else does.

khess

Good points and they’re well taken. My point is though, that I believe that web-based applications, including the OS itself, are the future because of their absolute scalability. How many virtual desktops can you service with a VDI vs. how many you can service if it were all web-based?

The zeitgeist for web-based OSs may as yet not arrived but it will and when it does, we’ll all laugh at VDI as old, clunky, and “can you believe we did that.”

And we are also judging web-based OSs based on what we’re currently seeing. What if we had totally poo-poo’d computers when the TRS-80 was introduced or the VIC20? Remember what mobile phones used to look and feel like? Bricks with antennae.

What we need to do is do more with less. If I had a Dell Inspiron Mini 9, with its “Linux Mode”–that would be enough because it boots in 2 seconds, has a battery life of 4 hours or so in that mode and has a web browser. And the document I spent two hours writing will be there when I drop my MiniNote into the pool, spill my beer on it, or my battery dies.

And RDP is not a great long term answer for a protocol, that’s why Cisco is leveraging its WAAS into VMware VDI and why Qumranet developed SPICE for their VDI solution.
But even these will go by the way of the dinosaur in favor of the web. The web is the only thing that will scale to a worldwide audience.

get_the_fish

“Current virtualized desktop solutions have the life expectancy of a fruit fly.” is simply wrong. “web-based applications, including the OS itself, are the future because of their absolute scalability.” flat out doesnt matter.

IT people mistakenly assume that the needs of the enterprise data center always outweigh, and can always out-muscle, the PREFERENCE of the user. Perfect example- Congressional members are demanding iPhones when the standard is Blackberry. We all know that from an IT perspective this is ludicrous and even dangerous. The point is that the don’t care about what we want. This is happening in enterprises everywhere.

The most important factor in VDI implementation is WHEN to use VDI, when to use centralized applications on standard desktops, etc etc. To reasonably expect that the “desktop experience” of VDI will be as rich as that of a native OS on physical hardware is foolish to say the least. However, for some users, the “richness” of the desktop is not only unnecessary, it is counter-productive.

For example, I have a mixed environment of virtual desktops (RDP) for a specific class of users with a specific job. These desktops are “stripped down”, lower resolution, no audio, no flash players in browsers, etc. This is by design- these users have no need for these features and management doesnt want them to have them. They are time wasters, plain and simple.

For some of the other job classes, where a rich desktop experience is a requirement, I have full machines with desktop OS’s in the traditional sense. The only real difference here is that these users demand this, as they could perform their tasks quite well in a VDI environment.

The argument made in this article hinges on one very flawed paragraph-

“Most users don’t care which operating system they use and many of them don’t know what they’re using anyway. What they really want is applications and more than 90% of all users use; a word processor, a spreadsheet, a web browser, and email. Why not remove the OS question completely by just using a browser for everything—even the Desktop itself?”

This is simply untrue. I don’t care what industry research pundits came up with this stuff. I can tell you- from my 15 years of experience as a network manager, then director, now CIO- users that actually use their desktops for something more than data entry into web forms, CARE VERY MUCH. I have seen more than a dozen desktop migrations fail because a handful of BA’s and execs dont like open office, dont like this or dont like that. It’s the simple truth.

With all these web based email apps, such as yahoo, gmail, Exchange’s web outlook, etc, why do most users in the enterprise still use Outlook? It’s a perfect example.

People seem to forget what got Microsoft into the data center in the first place- the fact that users were using Windows on their home PC’s, then going to work and expecting the same experience. People started to realize that it saves them a fortune not having to train users on their desktop OS when they are already familiar with Microsoft.

So, web based applications, web desktops, cloud computing, etc, all have their place, sure. Just not everywhere. Scalability is important, but none of that matters when the senior VP’s are firing their IT guy because they are sick of their computers not being able to play PokerStars (and you if you think that I’m joking on that, you have a lot to learn).

Wonderful paintings! That is the type of info that should be shared across the net. Shame on the search engines for now not positioning this post higher! Come on over and seek advice from my web site . Thanks =)

In this great pattern of things you’ll receive an A+ just for effort and hard work. Where exactly you lost me personally was first in all the details. You know, it is said, details make or break the argument.. And that could not be more accurate right here. Having said that, permit me say to you what exactly did give good results. Your article (parts of it) is rather powerful which is most likely the reason why I am taking the effort to comment. I do not make it a regular habit of doing that. Next, although I can certainly see a jumps in reason you come up with, I am not necessarily confident of how you appear to connect your ideas which in turn make the actual conclusion. For now I will, no doubt yield to your position but hope in the future you connect the facts much better.

I think this article is very helpful for us,it has solved our problem,thanks!
how much are real hermes belts http://bionor.es/cas/pdf/88/67/

Leave a Reply