Desktop Of The Future… Proposal
To begin, I feel like a quick homage to Conan O’Brien is in order:
Now that we have that taken care of, it is time to move on to the meat of this posting… specifically the “Desktop Of The Future”. [Note: every time I think of the “Desktop Of The Future”, I cannot help but think of the “In The Year 2000” bits on the Conan O’Brien show.]
I have this image in my head of what the future holds for my company. We have a fairly diverse set of functions we provide, we are geographically diverse, and use a couple of fairly data intensive applications for some functions. With such a distributed environment, management becomes such a nightmare. Combine that with the small number of people in my department and you have one painful looking headache to deal with.
End user technology has progressed at such a rate that we have a crazy amount of tools available to us. No longer are we restricted to using desktops with thick clients or web based applications solely. While those solutions provide some simplicity in deployment and management, the scale of which they need to be managed can be the choke point for IT departments. Instead, we have a whole new toy store that we can go shopping in to help determine how to keep users working properly. Admit it… we all like new toys.
There is no question that I have a penchant towards VMware (and virtualization, in general). So, I tend to lead more towards those solutions in the first place. However, their offerings are close to complete. Throw in a neat Microsoft product, and I believe we are well on track for what I would consider to be our Desktop Of The Future.
Problem: Determining how the desktop is going to be presented to the end user is the first step for us. Not necessarily the box sitting on their desk, but the picture of their puppy or Grandchild as the background, the ever-so comforting Start Menu, and My Computer (or “Computer” now). We have stuck to the typical workstation for those users that sit at a desk all day and laptops for those that “travel” (don’t get me started on traveling laptops). That has served us well since the beginning of time (or epoch, which ever came first). However, that is what is causing some of the management issues.. the sheer number of these devices on the network.
Proposal: Virtual Desktops ala View 4.5.
Reasoning: People are comforted when they have the “usual” experience. Believe it or not, people have a fairly emotional tie to their work environment. They spend much of their lives sitting in a little cube, doing whatever it is that they do. That little cube is their world. It has their plants, their pictures, and their personality. So, it also has their computer, their background, and their current working experience. Coming up with a remote desktop environment where users cannot even change their background is a touch dominating and will result in some sour-feelings and complaints towards the IT department. No one likes that, do they? I am sure I do not.
Theory: Using View 4.5, with the requisite vSphere environment behind it, we will be able to supply the users with that same/comforting style environment. They can click their Start menu to begin applications, they can change their background, etc… But, their computer will be sitting in the data center instead. Having the users closer to the data also means that their data-intensive applications will perform faster. Also, the quality of the hardware and the virtualization environment supporting View will increase performance. Management becomes easier in the long term… once those responsible for managing desktops make it over the View learning curve (linked-clones, single image, update/upgrade procedures, persistent desktops, checking out, etc…).
Plus, the ability to check-out a desktop image for a traveling user helps ensure more users are using the virtual desktops more often and everyone can benefit from the virtual environment regardless of the form factor of their device.
Application Presentation – Local Applications
Problem: Users need to use applications to get their job done. Otherwise, they are sitting at their desk doing nothing and collecting a paycheck (sounds kind of nice some days, right?!). Currently, we are using a hybrid environment consisting of some users using Citrix to access their applications while others are using locally installed applications. Believe it or not, users are using this Citrix vs. Locally Installed as a sign of dominance. People are actually thinking they are “higher up” if they have locally installed Office. In reality, they were our exceptions when deploying Office (for whatever reason), but that has been forgotten.
Additionally, similar to the desktop issue, the sheer number of deployments of applications crossed with the number of applications makes deployment and management makes things so much more difficult.
Reasoning: Install and configure one time, deploy to many.
Theory: Using ThinApp, we are able to provide the appearance of locally installed applications despite having never installed the applications on their desktop environment. Much like any new technology, training and learning curve must be overcome. However, once those skills are up to date, the flexibility of deployment options, application management, application security, etc… is immense. Additionally, using virtualized applications actually means that the desktop images deployed and used by end users can be more generic. Group membership and security determines which applications can be used, not the list of installed applications on the deployed image.
Additionally, we have 1-2 apps that are still running on… wait for it… Access ‘97. Seriously… it sucks. They are keeping us back from deploying more projects because of that requirement. Virtualizing that application would be a home run. If we never did anything from this posting except for virtualize Access ‘97, it would be a red-letter year!
Application Presentation – Remote
Problem: We have a number of applications that need to be presented via an RDP style service. We have been using Citrix to present these applications for all users, regardless of whether or not they use Citrix for Office or not. Additionally, we have a large number of external users that need access to a couple of these applications but we DO NOT want them to use a desktop environment internal to our systems.
Proposal: Microsoft Remote Desktop Services – RemoteApp
Reasoning: Cost effective as the product is included with Server 2008 R2 licensure (excluding the necessary CALs). Also utilizes advanced RDP protocol functions to optimize the network conversations between the server and the client (yes, I know that the client and the server would be on the same LAN, so optimized network conversations may not have a major impact).
Theory: These applications still need to be presented to our users. No questions asked. Done and Done. However, does it make sense to continue using Citrix services that become more and more difficult to manage, update, etc… or does it make more sense to move to a more simpler setup that provides the same level of functionality at a fraction of the cost? Every company has a budget… being able to cut the Citrix piece out for something that appears to provide the same level of functionality from a reliable vendor is a pretty easy pill for the business to swallow. Plus, as we use more and more Microsoft services, it makes sense as they all appear to build on one another.
I have a nice ThinApp and RemoteApp demo setup on my desktop and I have run it by a number of people in my organization. All in all, they really like it. They like the functionality and flexibility that ThinApp and RemoteApp provide. Ease of management, abstraction of the desktops to a very generic form, and the sheer WOW factor are the most positive comments I have received so far. WOW is pretty big around here. That is for sure.
Since my development environment is my desktop, I am looking forward to spinning up a little View environment on my machine to throw that into the mix. Then, I will have some ammunition for why I want to do what I want to do… live.
So far, everything has been going really well. I am crossing my fingers that I will be able to sneak some of this into the budget for next year.
But, even if this never turns into anything, or we select a subset of the proposal, the fact of the matter is that technology exists to change the working environment for our users in a very positive and exciting way. These technologies are not going anywhere. The competition is ensuring that. Instead, they demand to be refined and dropped in price until they are adopted in some fashion in the future. Getting in on the ground floor for these paradigms ensures that we are doing everything possible to ensure that our users can work effectively and efficiently. While most studies show that it is almost a wash when moving from a physical desktop environment to a virtual desktop environment… there is no accounting for WOW!