Citrix plans to enter VDI space: An analysis of their solution

Since my last article on the Virtual Desktop Infrastructure (VDI) topic, I have received a number of e-mails. Some show interesting tips, some are just angry at life (and me specifically at the time of their writing), and some point out some interesting solutions that are trying to fill the 'pie in the sky' VDI solution I discussed last time.

Since my last article on the Virtual Desktop Infrastructure (VDI) topic, I have received a number of e-mails. Some show interesting tips, some are just angry at life (and me specifically at the time of their writing), and some point out some interesting solutions that are trying to fill the ‘pie in the sky’ VDI solution I discussed last time.

Now personally, I haven’t seen a solution that makes me jump up and down saying “this is the end all be all for VDI!” However, I have seen some interesting takes on the problems I pointed out previously. Because of the e-mails and calls I received, I’ve decided to do a series of articles on the different VDI solutions available today. In this article I’ll discuss a solution (forthcoming) from Citrix for VDI implementations, and in future articles I’ll look at packages from Leostream, Propero, and any others I can get my hands on.

Before I get into the Citrix solution I need to qualify it some. First of all, the Citrix solution is not on the street yet (while the ones from Propero and Leostream are). Second, the Citrix solution I’m going to describe is a “stopgap” solution. Citrix is working on a full featured solution for release at some point in the future and is targeting a solution that covers most of what I described in the first article. The solution they are going to release now is a stopgap to help their existing customers with the VDI issues they face and is not meant as the end-game.

Okay, with the qualifications out of the way let’s talk about Citrix’s stopgap solution they’re calling the “Remote Desktop Broker” (or RDB). The concept with RDB is pretty simple; it’s an application that can feed parameters into the RDP client and provide you a way to manage connections and create resource pools of desktops (VMs, blades, etc.). The RDB application is installed on a Citrix Presentation Server and then published as an application. Users execute the application which then connects them to the type of desktop they (or the app) is configured for.

I’ll get into more detail on HOW it works in a second, but for now understand that it’s an application, not a server, and because of this is it uses a double-hop scenario with ICA connecting to the Presentation Server and then RDP connecting from the Presentation Server to the virtual desktop.

If you imagine an exiting Citrix environment (shown above) using an ICA client to connect to a Presentation Server, you’ll note that you have some of the items you need to provide a really solid VDI solution. You have a web interface that will allow for “publishing of the desktop,” you have the inherent SSL connectivity for remote users that Citrix already provides, and you have some session management and some tools to show connections and their status. So what’s missing? A way to pool desktop resources, handle peripherals, etc. Citrix’s RDB will handle some of this for you.

Citrix’s RDB will sit on top of your existing Citrix infrastructure. The concept is that the desktop broker application will be installed on your Presentation Servers. This application will communicate with an RDB database that will contain connection information for different pooled desktops. As pooled resources are used, they will be noted in this database so that the desktop is noted as “in use” and the next user will be routed to an unused desktop. If you look at the image above you’ll see that a connection will look something like this:

  • User connects to a  Citrix Web Interface server (or uses some other Citrix client)
  • Based on the credentials, a published application is made available for that user (in this case the RDB application).
  • The user launches the RDB application and is connected to the Presentation Server via ICA
  • Once the application loads, the user is routed to virtual desktop based on information in the RDB application. This is basically passing parameters to the Microsoft RDP client on the Presentation Server to tell it what end desktop to connect to.
    • During execution RDB is going to check its database to determine which desktop is available from the pool the user is connecting to.
    • This information is passed to the RDB application which in-turn uses it to establish an RDP session to the Windows XP VM or blade.
    • When the application is closed the desktop is released.

In looking at this you’ll notice right off that be that this is a “one session for the price of two sessions” deal. It’s a double hop. While it gets you to the desktop, its obviously not the most optimal configuration around. Citrix knows this, and this isn’t their long term design, but with this should come some cautions.

  • Speed screen latency reduction features will not help here. Most functionality/assistance you get with speed screen latency reduction features focus on making the users think that latency is not effecting their ICA session. Basically it works by detecting the font and size and color of the text being typed in the Citrix session then instead of waiting for that text to appear on the client it shows it on the client before the round trip communication ever happens. In this case the RDP client is essentially an image or movie like application and typing within it is an image change, NOT text within a field or form on the Citrix server. (Check out this article for more details.)
  • Not all features available in Citrix will be available in the RDP session. There will be some features (noted later in this article) that you are used to with Citrix, but since it is not ICA to the XP Pro, it will not be full featured.

I could put a third point here about it being a double hop, and the cost of a Citrix license without getting full features, etc., but I think that’s implied with these two other cautions.

In caution number two I noted that you won’t get all of the features (Speed screen, good printer redirects, driver replication, pure session mgmt in the XP pro desktop etc), but you will get some. Citrix plans to offer the following:

  • Client Drives Access. Users can share drive mappings between their hosted desktop and server drives. I must note I had mixed results with Client drives mapped to the Citrix session then mapped to the XP Pro session, but server drives did show up in my XP Pro desktop.
  • USB Key Drives. Users can access USB devices through their hosted desktops; however, the key must be connected prior to session startup.
  • Audio. Users launching a hosted desktop the user can hear audio (for example, using Microsoft Media Player), though this may require some ICA channel reconfiguration if you use RDB on an existing application host.
  • Single Sign On. When the user logs into the Citrix Web Interface and connects to the published hosted desktop, they can be automatically logged into their virtual desktop using Citrix Password Manager. This one was kind of a pain for me as I do NOT use password manager and I was required to sign into the Web Interface and then Sign in again to the Virtual Desktop. (Suggestion: Password Manager for free would be nice!)
  • Printing. Users can print to locally installed printers as well as network printers. I had a couple of interesting issues with this that I believe are more related to the XP Pro desktop than the Citrix session. Again it leads back to not having a control agent with the XP Pro desktop and needing to do some manual configuration on the images used by the users.

The long and short of this solution is that it is a stop-gap. Basically it’s an application that will allow you to create pools of VMs or XP blades for your users, and use your Citrix servers as a proxy mechanism. Citrix is hoping that it can help its existing customers get over the big hurdles here by delivering a brokering mechanism and allowing them to leverage the existing Citrix tools to solve the other issues I’ve previously mentioned (secure remote access, web-based authentication, etc.).

The disappointment to me is that Citrix wasn’t out in front of this a long time ago. Last year at VMworld people were all over themselves talking about Citrix servers on VMs or about running VM workstation instances on Citrix (which, by the way, is a ridiculously stupid idea) instead of focusing on creating real VDI tools and solutions. Hopefully the time-to-market on this won’t be too long and the follow-on product that will solve most of this product’s issues will not be to long of a wait.

In my next article in the series I will take a look at Propero’s VDI solution.  I was going to use Leostream’s solution first but had a slight issue with it during configuration that I will have to discuss during the article. As always, e-mail or post comments, I look forward to most of them. J

Join the conversation

38 comments

Send me notifications when other members comment.

By submitting you agree to receive email from TechTarget and its partners. If you reside outside of the United States, you consent to having your personal data transferred to and processed in the United States. Privacy

Please create a username to comment.

Nice article Ron. No ETA on the RDB right?

You probably wrote this article before Wyse announced their VDI product. It seems like everybody's racing for VDI.
You should do an article on the Wyse Enterprise Desktop Virtualization Solution.


You will right? ( no pressure here )

Regards,
Michel.
Cancel
Good article. To enhance this series, I would love to read your analysis as to WHY a business would go to this model. It certainly does not seem for everyone. Citrix/TS has been doing virtual desktops since their beginnings. It appears that in this 'stopgap' solution, the user will still need to distinguish between their real desktop and their virtual one. Users don't get so confused about it like they did 10 years ago, but a lot can be said for just having one(1) Start Menu button.
It appears that this 'stopgap' will still have a 'real' full OS on the client with another virtual FULL OS running on top. I think Citrix would do well not to waste time with this and put more resources into extending their Virtual APPLICATION Infrastructure to push/pull self contained apps to the desktop. That would truly reduce management costs by allowing the desktop OS to be more locked down without hindering useability of the virtual app. Yes, I know Softricity already does this...
Cancel
I hope citrix fixes some of the issues with RDP over ICA.
-Active X client double click issues.  Double clicks are not registered
-Java client icon issues and keyboard issues.
-Disconnect issues.  Disconnect from ICA session but RDP session still running.
 
 
Cancel
We have been using this double-hop method for some remote users running the RDP client on presentation server with the users connecting to that over ICA and have observed the following:
 
- When users do anything that causes alot of screen updates, ICA session bandwidth is 2-3times more than if the same operation is performed over native ICA. We've had to cap ICA bandwidth via policy as a single user could easily saturate their link.
 
- We get very bad performance over very latent links (no speedscreen and higher bandwidth) whereas native ICA performs fine (even without speedscreen enabled!).
 
I suspect the RDP client is just screen-scraped in this double-hop scenario, which means none of the optimisations that ICA perform on the GDI calls will get used - it will just be sending a compressed bitmap'ed image of the RDP screen down to the ICA client.
 
I can see people running into lots of similar issues if they deploy this and expect performance to by anywhere near what they are used to running native ICA on presentation server. Make sure you pilot extensively!
 
Cheers,
 
Neil
Cancel
Is it me, or would you still be limited to 20-30 users per Citrix CPU?  If that's the case, you're not really getting a good ROI on anything since you'd need a big Citrix environment to address a medium to large organization's VDI solution.  
 
I could see some value in VDI if an org replaced the phisical desktops with thin clients and accessed VMs using RDP, or the flip side, a solution that leveraged the existing desktop resources without the overhead of a terminal server proxying all that data.  Possibly something like a Route1 MobiKey, but instead of a one to one relationship between the client and the desktop, have a dynamic desktop resource pool that can be shared by all users, possibly in conjunction with a product like Softgrid so all desktops can be provisioned for any application the user needs access to upon login.
 
This VDI solution seems like it would add a lot of overhead for nothing... assuming the User/CPU ratio limitation of a standard Citrix server.
Cancel
I don't know, I think this is starting to over complicate things..  Is it me or is Citrix not already a VDI solution a heart?  Really why bother..  If starting from scratch, you could provision thin clients to everyone and simply use Citrix for a centrally managed desktop infrastructure.  Who cares if the desktop is a VM or a terminal instance of a desktop, it's still a desktop.  Citrix already has a wealth of connectivity options so you don't need to address getting access to your desktop from anywhere.  This seems to me like a perpetual motion analogy, i.e. make things as complicated as you can so nobody notices the inhearent flaw in the design, until they try and get it to work...  VDI at this point is basically a solution looking for a problem, not vice versa...  There are plenty of solutions out there to address your problems, do we really need another promise that won't deliver?
Cancel
remote developers (here or in india) need access to a full standard-image developer desktop.  In a large company like ours the desktop team has a standard developer image which would be a pain to retrofit into the Citrix environment.  just give them RDP and put a developer desktop pc somewhere- they're ready to go. 
people who work remotely sometimes and want full access to their regular desktop (we have about 1000 concurrent workers on saturday mornings (currenlty they launch pub apps, i'm sure they'd prefer a fullscreen of their regular desktop)
Alternative to server-based remote seasonal capacity (e.g. add a bunch of claims adjusters with full desktops during hurricane season)
Pandemic insurance (e.g. a region is quarantined and people can't go into the office and the company is still fully operational)Me wanting to continue work from home and pick up where i left off (with about 25 different windows open)

We have server images that are built by a server team and desktop images that are built by a desktop team.  Unfortunately, we're in the middle as Citrix technologies.  We believe that all we do is provide virtual desktop machines, so to us a server is really 100 desktops, but we haven't been able to bridge the gap between the two teams.  So getting some desktop stuff to work on the servers takes some effort.  different packages, different SMS distro, have to start a project, test, buy servers, test more.  if a team of 20 people says they want to work remotely occasionally, we say "Ok, go ahead" and there's nothign that has to be done (we don't care what apps they use or whether or not we already support them on citrix) except to tell them to remember their computer name - which a VDI solution would automate.

From my experience and in those examples I think VDI will have some useful niches.

some consultant who used to be onsite at our company said eventually desktop computing will be a grid and you'll just be brokering memory and CPU from other machines.  this is kind of a step in that evolutionary direction.  i'm coming in, i'm saying i want to run a desktop, and a service is brokering for me where that provider will be.  maybe it is solution before a problem and like a previous post said it will change again by the time everyone figures out how to do it and i guess that's one of the ways we progress.
Cancel
I still think app virtualization will outpace desktop virtualization and address all the VDI requirements using SBC/Citrix.  The average home user will never be interested in VDI, since set-top devices really haven't caught on much anyway and we're all box-huggers who like powerful video cards and lots of ram, but they'll definately want app virtualization to isolate them from the OS.  I would rather see advances in app virtualization like full context support (comunication between virtualized application suites) and support for any drivers and specific services.  That way you can have the best of both worlds without having to totally rip and replace everything you've invested in thus far.  I think MS and Softricity will come up with some very appealing solutions to address the SBC shortcomings before VDI claims any significant market place.  It just seems more logical.
Cancel
Ron,
Just some thoughts...
Thanks for the analysis/information. I think we're all kind of looking at niche possibilities and making them a bit bigger than they are. The headlines get our attention (when a thin client vendor joins the VMWare VDI club, people listen), but what are they saying?  From a resource and complexity viewpoint, VDI is less efficient and cost-effective than SBC. Let's look at the backend resources,and ask questions. How much disk space does that VM need? How much RAM? And where is that diskspace? The last time I checked, SAN space was still more expensive than DASD. 
And the connector? Let's hope that RDP gets less latency-sensitive.
I think our collective problem is the pervasive focus on "the box".
Don't look now, but "the box" and it's collection of discrete software chunks are mutating.
It's STILL about the app...

 
Cancel
but I digress, Ron in reading your article I was really confused on what Citrix is trying to accomplish.
They already have what you would call a VDI, and have had it for years.  What they are trying to do
confuses me from the standpoint, of what does it actually accomplish for them. So they build a broker to
to VM Workstations, doesn't this dilute the benefit of PS 4.0.  I believe it does.  THey can tought VMWare as
strong partner for this, but I truly believe VMWare will push Citrix aside.  If I have a thin client, or desktop, and I can RDP into a VM, please tell me why I need Citrix.  I have a desktop, I have VD, what is Citrix exactly bringing to me.  It is typical Citrix in the Johhny come lately, me to bandwagon, with a half baked solution.  THey should be defining this space, not figuring out what VMware is going to do.
Cancel
ORIGINAL: Yonderbox

Ron,
Just some thoughts...
Thanks for the analysis/information. I think we're all kind of looking at niche possibilities and making them a bit bigger than they are. The headlines get our attention (when a thin client vendor joins the VMWare VDI club, people listen), but what are they saying?  From a resource and complexity viewpoint, VDI is less efficient and cost-effective than SBC. Let's look at the backend resources,and ask questions. How much disk space does that VM need? How much RAM? And where is that diskspace? The last time I checked, SAN space was still more expensive than DASD. 
And the connector? Let's hope that RDP gets less latency-sensitive.
I think our collective problem is the pervasive focus on "the box".
Don't look now, but "the box" and it's collection of discrete software chunks are mutating.
It's STILL about the app...



 
No doubt. If you had been in any of the sessions I did I briforum I stomped around and basically sounded like the biggest anti-VDI guy around (at least to some people). When the whole time I was jsut trying to say that unless you had specific needs for isolated, unique desktops there was no business driver (read cost savings) to goto VDI over SBC or other app deployment models.
 
Dont get me wrong I think it has its place. And being a Citrix, app virtualization and Vmware guy I think I have found myself in a great postion to objectivly (and with experience) talk to the potential solutions and when they fit an dont. Should be a good series of articles.
 
Ron
Cancel
ORIGINAL: Guest

but I digress, Ron in reading your article I was really confused on what Citrix is trying to accomplish.
They already have what you would call a VDI, and have had it for years.  What they are trying to do
confuses me from the standpoint, of what does it actually accomplish for them. So they build a broker to
to VM Workstations, doesn't this dilute the benefit of PS 4.0.  I believe it does.  THey can tought VMWare as
strong partner for this, but I truly believe VMWare will push Citrix aside.  If I have a thin client, or desktop, and I can RDP into a VM, please tell me why I need Citrix.  I have a desktop, I have VD, what is Citrix exactly bringing to me.  It is typical Citrix in the Johhny come lately, me to bandwagon, with a half baked solution.  THey should be defining this space, not figuring out what VMware is going to do.

 
Really they have been messing with this concept for a year...  What they are really trying to do is keep their existing customers happy with a stop gap, until their full blown direct connect solution comes out, and keep them from going to other vendors to fill a need and thus risk loosing those customers.
 
From a why do I need Citrix perpective?  Well, read that first article of all the missing things in a simple thin client and XP pro desktop environment and you will see why all these software vendors are jumping on the bandwagon. Basically thign client to XP pro works if you have 10 or so clients. if you ahve 100 you have more on your plate and someone is going to write the code to make simple, easy to manage, load balanced, centrally configured, etc etc. Citrix jsut wants that action and if they can get a product out there fast enough they can get it to their existing client base before they buy another solution.
 
Ron 
Cancel
ORIGINAL: Aiian

Good article. To enhance this series, I would love to read your analysis as to WHY a business would go to this model. It certainly does not seem for everyone. Citrix/TS has been doing virtual desktops since their beginnings. It appears that in this 'stopgap' solution, the user will still need to distinguish between their real desktop and their virtual one. Users don't get so confused about it like they did 10 years ago, but a lot can be said for just having one(1) Start Menu button.
It appears that this 'stopgap' will still have a 'real' full OS on the client with another virtual FULL OS running on top. I think Citrix would do well not to waste time with this and put more resources into extending their Virtual APPLICATION Infrastructure to push/pull self contained apps to the desktop. That would truly reduce management costs by allowing the desktop OS to be more locked down without hindering useability of the virtual app. Yes, I know Softricity already does this...

 
No doubt. but for multiple reasons business will use VDI. I see it running next to TS servers in the long wrong. Security would be one reason. a true need for desktop control by power users that have different apps could be another, centralization of unique desktops for DR, mgmt, recovery etc is another. Its a tool. Just like App Virtualization. I have had IT people int he alst couple of years tell me I was crazy for thinking app virtualization was a big deal.. They said "we just create silos or add servers or install the app on the desktop, why mess with app virtualization???"  Of course app virtualization isnt needed everywhere. but it is a valuable tool, just more valuable to some orgs than others.
 
Ron
Cancel
Depends. Overhead I have seen in what I am running is minimal. I mean you have a session, and the RDP client running. On one of todays boxes you could get 50-60 or more users on a daul pretty easy. The trick I see is the licensing of the Citrix licenses to support this model. I mean a dual proc server can be picked up for 3 or 4K but 60 Citrix cals at 300 or350 a pop is 20K or so...
 
hardware is a commodity. Software is king, and we pay the tax to the king.
 
Ron
Cancel
It is. Check out the original VDI article I wrote and we discuss why someone would use one over the other. TS is also a by far CHEAPER solution. but VDI fills a niche.
Cancel
view of Citrix, in the past which they defined a market to present. I believe that the PS business is in for
a rude awaking to due to many factors. First of all with the recent announcement of Citrix taking SA direct, they will
alienate partners, not all but quite a bit.  The channel is their extended sales force. They will lose some good partners.  Secondly, with Longhorn, will that kill Citrix in the enterprise most likely not, but in the small to medium business, with 3-10 TS servers, why exactly do I need the value of Citrix, and if you talk to a Citrix rep, they can't quite explain it either.  Third of all, VMware, will now be a competitor of the SBC.  I have seen several large scale deals first hand both choose hosted XP over RDP over Citrix.  VMware owns the Server Virtualization, but to grow and expand, you have to tackle the desktop, and they will position this again and again.  Fourth of All, with the recent buy of Softricity by Microsoft, this totally cuts Citrix at the knees with Tarpon.  I can tell you first hand that this way a big push by Citrix, much easier to beat up a small company from Boston, much harder to fight the hand that feeds you.  Microsoft wil fix app problems on TS, but the bigger issue for Citrix is if I am not a remote site, or connecting remotely, why again do i need Citrix to manage my internal apps.  I may be missing something, but the dys of being the only game in town are done....
Cancel
So, this can be accomplished already by publishing an RDP client with an RDP file that points to your desktop and doing this per user.  So, basically, a whole lot of nothing.
Cancel
Don't hold your breathe, it's going to be tough enough for them to develop launching the RDP client!
Cancel
Why bother, I'd expect to see the next iteration of this offering to have ICA direct from the XP (Vista) Image, with all the benefits an ICA session provides.
 
Add a whole host of management features over the top of that and it becomes a real strong offering to those areas where VDI is suited.
 
Place your bets for Q4 07 (Global iForum launch???)
Cancel
I have to say, for an extremely niche solution there's a lot of talk about VDI these days as if any business can make use of it.
 
In my view about 90% of the time a business can't exclusively work in an SBC environment it's because of either using graphic-intensive applications of some kind, or needing to work offline. Neither of which VDI solves. (I'd be very interested to hear other people's feedback on this.) Not-easily-solvable App compatibility issues i haven't encountered for years, particularly using isolation or Softricity.
 
To me that makes VDI attractive for such a minute slice of businesses, i just don't quite get what the fuss is about.
 
I grant you, the scenario of odd-ball apps needing to be run by one or two staff is absolutely more complex and timeconsuming in an SBC environment, but still not as expensive as needing to essentially buy, build, manage and maintain 2 computers for them.
 
On the other hand, i have to agree that requests such as staff being able to connect to their PC in the office during a pandemic are quite common and VDI does solve that... oh hang on, so does SBC. (i'm sorry about that, but that has to be the funniest selling point for ANY solution i have EVER read - and i've read a few...)
Cancel
I think you pretty much summarize the "general" Citrix viewpoint on this solution, and why we haven't done very much to address this virtual desktop model.  As Ron stated, Citrix didn't move very quickly to provide a solution for this; instead we kept focus on enhancing Pres Server (i.e. building the 64-bit version) and building new technologies like Tarpon (which actually has very interesting applicability to a virtual desktop scenario).   We felt that Pres Server was already a virtual desktop solution--so why should we embrace this model of virtual WinXP desktops?
 
Well, in an nutshell we've found that there are two good reasons for done a virtual desktop instead of TS+Pres Server
 
1. You have lots of apps and want to reuse an existing desktop image instead of migrating everything from WinXP to Terminal Services.  With a small number of apps, this isn't a big deal to do the migration; but with lots of apps it can be  time-consuming and costly.  Tarpon can actually be introduced into a scenario like this--just as we've seen AIE and Softgrid complement TS and Pres Server environments. 
 
2. The "remote developer" scenario.  This just isn't the greatest fit for TS+Pres Server for a variety of reasons.
 
VDI and hosted desktops are likely to remain a niche solution for some time.  TS / Pres server should remain the predominant way to "virtualize" apps and desktops.  But, as I've heard Ron wisely say: "As a consultant you always want to have options"--and this is one more option to have in your toolbox.
 
Bill Carovano (a Citrix employee)
Cancel
ORIGINAL: Guest

Why bother, I'd expect to see the next iteration of this offering to have ICA direct from the XP (Vista) Image, with all the benefits an ICA session provides.

Add a whole host of management features over the top of that and it becomes a real strong offering to those areas where VDI is suited.

Place your bets for Q4 07 (Global iForum launch???)

 
 
How about enhancing RDP with latency reduction, JPEG compression, and 3D OpenGL support, among other things? If and when this happens, what advantages will ICA offer then?
 
Peter Ghostine
Provision Networks
www.provisionnetworks.com
Cancel
Hi Bill,
thanks for the reply. I'm glad that Citrix see things in a similar way on this issue.
I still agree that VDI has it's uses for the (very) occasional situation, and as such being able to install a PS for winXP with an ICA listener, along with all the other bits that make it easy not to have to touch an individual ctx server would be very useful. But if i compare that to other Citrix projects such as Ocelot (that's the opengl one, isn't it? ), Tarpon and even IRIS, it simply is not that significant.
 
Not to keep going on too much, but i've had far more need for IRIS-like-tech than a VDI solution.
An worst-comes-to-worst, there's always goto-my-pc... lol.
Cancel
Virtualization
Cancel
Virtualization is a hot area of IT these past years and it seems every vendor between the water and heaven is jumping onto the bandwagon.  While the concept of virtualization is great, from my perspective, I feel very few vendors have a "REAL" value proposition in this field.  Most are "virtualized" just for marketing sake but lack a real technology solution.  I can be wrong, but my gut feel with VDI is a prime example of taking a "Real" mess and turn it into a "virtual" mess.  Either way, it is still a mess.  And there are several reasons why I believe VDI is a not a good solution.
 
1.  An extra layer - PS already adds an extra layer of complexity to how applications are processed and screens refreshed.  Now we had virtualization which on its own has its own layer of abstraction and overhead.  That has got to have some performance overhead.
 
2.  Less features - what is the business value of deploying a solution that has less features and potentially more headaches.
 
3.  Streaming vs hosted - Citrix has been the dominate leader in the world of application hosting.  However, I believe this is starting to change.  I believe the future is not in hosted application but rather in streamed applications.  The wide adoption of high-speed broadband, and will continue to increase, will further push technology such as Softricity to the forefront.  The real issue with Citrix solution is that it is a server centric solution.  While running standard office based apps are okay, running multimedia rich apps will present some major problems in performance.  Unfortunately, the world is moving towards multi-media rich applications.  This is where application streaming comes in.  The applications are stored centrally so it is easy to manage, deploy, and maintain, but it is streamed to the desktop so that fully uses the local processing and memory.  The next big step to compliment streaming application is streaming data.
 
The only vendors, and I am sure there are many that I have not seen yet, that I feel have a done a great job based on what I have seen is VMWare, Softricity, and Citrix PS product line.  Now when I say vendor that have done a great job, what I really mean is the technology solution these vendors bring forward.  Like server virtualization, application virtualization, and etc...  Sorry, I just don't see VDI being a real virtualization of anything other than a mess of virtualization.  But I understand why Citrix is jumping onto the bandwagon.  It is feeling the pressure of competition.  It is feeling the market is slowly sliding away right under his own feet.  When the acquisition of Softricity by Microsoft was announced, I knew then and there, the world of Citrix is at jeopardy.  Softricity with the Citrix PS product line could have been the de-facto solution that would have differentiated Citrix from MS TS.  Instead of installing Office or any applications on each Citrix Server, with Softricity, they could have had a complete solution that allowed customers to quickly deploy standard base image of Citrix server quickly and easily.  But now with Softricity in the hands of MS, the value of Citrix/MS partnership seems less attractive. 
 
Just my 2 cents...
Cancel
Just a counter 2 cents worth to this on the Softricity front which nobody seems to even acknowledge as a possiblity -
 
Softricity is just an app like any other developed by a bunch of smart people. Well Citrix have a bucketload of smart people too so then what's to say that Citrix doesn't develop their Tarpon into just as powerful a solution as Softricity. If that comes to pass, and MS incorp. Softricity into the seperately purchase SMS, then what's to say that people wouldn't buy Citrix with Tarpon streaming rather than Citrix and SMS or TS and SMS?
 
Let me just agree once more for clarity - Softricity is great, but still doesn't mean that Citrix can't come along and clone it, plus integrate it very deeply with PS, ICA channel and CAG for remote operation of streamed apps, etc.
 
Hosted apps will never go away, i don't think having _just_ streamed apps is a viable solution for many situations. For starters if you have any remote sites it would suck severely to suddenly have to set up replicated db's, fileshares, etc to accomodate those remotely running apps.
Cancel
I agree..  For the foreseeable future, hosted app won't go away.  And the reason is like you said, data and file share is still a problem.  What citrix and softricity have done very well in is in application delivery.  Whether it is hosted or streamed, it does not matter.  The important thing is that they have found an innovative way to deliver application.  On the other hand, what about the actual data.  Still the same problem.  Need high-speed bandwidth or else performance sucks when transferring large file.  Hence I said Softricity with data streaming technology is a killer combo. 
 
There is a couple reason why I don't see Tarpon being a competing product with Softricity.  Again, these are just my gut feelings and nothing set in stones...
 
1.  Softricity is light years ahead of Tarpon.  It is proven product in the field with real life deployment.  They have real-life field experience already.
2.  Tarpon is what I call a catch-up product.  That means Tarpon was developed to offer a similar product to Softricity so Citrix can better keep their customers.  From a business perspective, this a defensive move vs an offensive move.
3.  Microsoft and Softricity.  Microsoft is very good with developing catch up products.  IE, TS, the Windows OS.  Tonnes of example where Microsoft came late into the market with an innovative product but over time have caught up or totally eliminated the competition.  Microsoft can do a pretty successful job with catch up products, imagine what they can do with a proven product that already has market leadership, Softricity.  That is a scary proposition, and I was in Citrix position, I think I would say the fire is burning awfully close to my butt.
 
If you can compare the value proposition of Citrix 10 years ago vs now, I would say many people are starting to question the value of Citrix.  And by that I mean the value of PS product line.  For many, TS and Citrix are relatively the same.  Only businesses who require specific function of Citrix that TS does not offer at this time or technical people like us know the differences between TS and Citrix.  Even that, it is getting harder and harder to differentiate.  With TS Longhorn, the gap is narrowing and TS is gaining grounds particular at the SMB level who don't require fancy features. 
 
Now, I don't know the specific partnership arrangement of Microsoft and Citrix, but I would like to pose some questions for us to ponder?
 
1.  Why did Microsoft originall partnered with Citrix when MS has a competing product? 
2.  What will happens to Citrix if one day, MS TS does everything that Citrix PS does?  Today, MS TS can host apps, can publish apps, and with Softricity, can stream app.  What will be the value proposition of Citrix to the market and the MS/Citrix partnership?
 
 
We live in interesting times and I look forward to the outcome of some of the major plays these vendors have set into motion.  Only time and market reaction will tell who has made the right decision.  My bet is that Citrix will be under extreme pressure.  The growth to $4 billion company won't be the same as growth to $1 billion company.  And with fire under its main product line PS, will Citrix have the leadership to lead or to follow?
 
Cancel
ORIGINAL: Guest

Today, MS TS can host apps, can publish apps, and with Softricity, can stream app.  What will be the value proposition of Citrix to the market and the MS/Citrix partnership?

No it can't. MS have been promising these features for 3 years now! And the forthcoming 'published apps', the only feature they are halfway close to catching Citrix with, are *really* hokey.

Citrix will continue to be the value-add it has been ever since NT4 TS. Better performance, better management, better security, better end-user experience.
Cancel
Terminal Server in Longhorn is targeted to less-complex small to medium sized deployments, w/o the need of a 3rd party add-on like Citrix.  Large companies will continue to find value in Citrix Presentation Server, but also in competitive products like Provision Networks Management Frameworks Enterprise Edition and Ericom Powerterm WebConnect, Jetro CockpIT, 2X...

Longhorn still lacks a Terminal Server Load Balancer, Universal Printer, Memory Rebasing, Profile Management, TWAIN Redirection, bi-directional audio, bandwidth management, granular control of client resources, delegation of administrative tasks, Java Client..... so there's plenty of room left for other companies to add value on top of TS.
 
AFAIK MSFT has no intention of "catching-up with Citrix" in the near future, as there's no reason for them to do so.  Citrix Clients have to purchase TSCALs and Server Licenses, so MSFT prospers w/o threatening Citrix.

http://www.sessioncomputing.com/add-on.htm#suites
Cancel
It is correct, TS is targetted for small and medium based organizations that require less-complex setup.  Unfortunately, as of today, majority of the business in the market today are small medium organizations who are only looking for standard or basic setup.  The question is not whether Citrix still has value today.  The question is will Citrix still have the same value as Microsoft continue to improve TS and add more Citrix-like features and creep up the food chain and replace Citrix-based customers with their own TS/Softricity/SMS solution? 
 
The challenge with Value-Add is that it is simply that.  A value add.  An add on service.  It is not required.  If we compare TS and Citrix today, the core function of these two products from a business purpose perspective is identically the same.  They simply host applications.  That is it.  Everything after that is simply a matter of preference.  Let's do a simple math and maybe we can see the Microsoft strategy at work.
 
Citrix claim current count indicates that there are over 164,000 Citrix PS on the market today.  If you estimate each PS server services 140 users, that means Citrix has about 23 million end users in production worldwide.  That is a fairly impressive customer base.  Now let's look at MS TS customer base.  For 2005, according to IDC, 30% of ALL servers sold were deployed with MS Windows or roughly 1.19 million MS Windows Server.  Having said that, it also means there are potentially 1.19 million TS servers in the market place today vs 164,000 PS Servers.  Now let us say that because TS is not as advanced as PS, each TS can only handle 50% of the load of PS.  That means MS would have a potential customer base of 83 million TS end users in production worldwide vs 23 million.  And let us not forget for each of the 23 million ICA user, customers also has to purchase TS license from Microsoft. 
 
Having outlined the above assumptions, let us say, Microsoft adds new Citrix like features or improve existing features into the mix over the next 3 years in its longhorn TS product line and future OS version.  Better RDP protocal, better web interface, better video and audio performance an handling, better management console, better load management, and etc, where do think Citrix will fit in.  We cannot deny the fact TS today is a much improved version of the original "Microsoft Windows NT - Terminal Server Edition".  So the question is not IF MS will add the features,  it is WHEN they will add the features.  If my gut feeling is right, features are added based more on strategic value than anything else.  On top of this, MS adds Softricity into the application delivery space enabling streaming and hosted application methods.  Where will Citrix be?  What is the value proposition be for the end-customer when TS can do 95% of what Citrix does? 
 
The challenge with Citrix is that it sits on top of Microsoft.  The issue with this model is that should Microsoft decide to one day not to renewal its licensing / partnership agreement with Citrix, where do you think PS product line will be?  Of course, I don't think Microsoft will do that today as it needs someone to fund the development of the TS product line.  Why fund it yourself when you have someone that is can will fund it?  If you have a great invention that require huge capital investment, would you spend your own money or rather have the funding of investors.  With each PS sold, users have to buy TS licenses from Microsoft which essentially means Citrix is funding MS TS development efforts at some level.  If I was Microsoft, of course I wouldn't severe the relationship with Citrix either.  Hell, Citrix spends the marketing dollars on selling PS and in return I get TS licensing revenue without extra effort.  No one in his right mind would severe this kind of relationship.  Of course, that is until one day I have developed and incorporated most of the Citrix like features into my own TS product line and have a customer base of 100 million users.  Why do I need Citrix then?  Of course that is if I run Microsoft today.  But we all know how business-friendly Microsoft, Bill and Steve, are so I wouldn't worry about it.
 
IBM OS/2 had a licensing agreement with Microsoft for Windows development for 10 years.  And we all know how well OS/2 did.  Last time I checked, I didn't see OS/2 as a supported OS.
 
 
 
Cancel
If MSFT wanted to they could:

A.  Purchase Citrix
B.  Hire ~ 100 employees and hammer out a feature set to rival Citrix PS

That being said, I don't think either of these is in the works (unless Apple decided to purchase Citrix), as MSFT makes a killing w/o doing either of these. 

P.S.  Citrix Servers hosting >=140 sessions are the minority, not the majority. I'd bet that ~ 2% of all Citrix Servers host this many concurrent sessions, and the average is around 50-75 concurrent sessions.  Only the Enterprise Edition of PS can host more sessions than the base TS product can, because it has built-in technology from Aurema & RTOSoft.  Citrix isn't usually needed to manage servers scaling up, but used to manage a farm as it scales out. 
Cancel
the short and sweet is this:
MS provide the bread, Citrix provides the butter. any company that wants a fancy buttered bread obviously still needs to buy the bread, so to say that MS are competing with Citrix... well no just simply not the case. Every Citrix customer is an MS customer.
 
In fact as per Windows EULA, any desktop / published app type connection to a windows box must be via RDP or go via RDP. (or something to that effect) Hence Citrix sessions consume a TS license + a Ctx license.
 
But at the same time, there's no reason for MS not to improve their base TS, just because Citrix might have to get off their bum and improve the PS experience even further. It would just be plain wierd for MS to leave their TS featureset at the Windows NT level, just because there's no direct competition. Not to mention all the TS-only customers who pour all their Software Assurance money into MS hoping for new features to come out of that.
They'd be pretty disappointed if MS said, we'll have your SA money thank you muchly, but if you like more features than what you have in this version, go and buy Citrix.
Cancel
Citrix uses the ICA Protocol, not RDP, but still uses the terminal services service.
 
"In fact as per Windows EULA, any desktop / published app type connection to a windows box must be via RDP or go via RDP. (or something to that effect) Hence Citrix sessions consume a TS license + a Ctx license."

 
ORIGINAL: Guest

the short and sweet is this:
MS provide the bread, Citrix provides the butter. any company that wants a fancy buttered bread obviously still needs to buy the bread, so to say that MS are competing with Citrix... well no just simply not the case. Every Citrix customer is an MS customer.

In fact as per Windows EULA, any desktop / published app type connection to a windows box must be via RDP or go via RDP. (or something to that effect) Hence Citrix sessions consume a TS license + a Ctx license.

But at the same time, there's no reason for MS not to improve their base TS, just because Citrix might have to get off their bum and improve the PS experience even further. It would just be plain wierd for MS to leave their TS featureset at the Windows NT level, just because there's no direct competition. Not to mention all the TS-only customers who pour all their Software Assurance money into MS hoping for new features to come out of that.
They'd be pretty disappointed if MS said, we'll have your SA money thank you muchly, but if you like more features than what you have in this version, go and buy Citrix.

Cancel
Take a look at the most recent posting about Microsoft reducing Softricity by as much as 85% list...  and take a look at the section that says:
 
In terms of Citrix Streaming Server (aka Tarpon), Softricity is now the first product that directly competes with Citrix. Presentation Server was always an "add on" to Terminal Server. However, customers will now have a decision to make between software virtualization platforms. It will be either Citrix OR Microsoft, but not both. I am curious to see how this will affect (if at all) the relationship between Microsoft and Citrix.
 
This is not quoted by me nor was this publicly available to me when I made the prediction in my previous email.  Do you see MS starting to turn the table now?  TS is the bread and Citrix is the butter.  For now.  But why would you just want the bread when you can have both?  Like you said, the TS team at Microsoft is prospering..  Great.  But Bill Gates and Ballmer don't just want prosper.  Understand their mentality.  Prosper is a "Low bandwidth" vocabulary.  'Domination" is more in line with their mentality.  They did not build MS from 0 to a $25 billion company on the basis of "prosper" as a goal.  MS is a long term company.  And by that I do not mean 3-5 years.  That is a blink of an eye.  They are 10-15 year investment company.  And the scary part about how MS works is that they will give you 5 years of "perceived" prosperity in exchange for 20 years of total domination.  That is how they gamble.  And with a $50 billion cash reserve, I would have to say MS is one hell of a gambler.
Cancel
I could be wrong, but it looks like that statement was made by Michael Burke, not by someone at Microsoft.  Perhaps someone will clarify if that was a quote from MSFT or an editorial from Michael.
 
"In terms of Citrix Streaming Server (aka Tarpon), Softricity is now the first product that directly competes with Citrix. Presentation Server was always an "add on" to Terminal Server. However, customers will now have a decision to make between software virtualization platforms. It will be either Citrix OR Microsoft, but not both. I am curious to see how this will affect (if at all) the relationship between Microsoft and Citrix."
Cancel
I doubt Microsoft will quote this directly to the world...  Of course it would not be a wise move...  The important issue is not who quote it... But the fact the someone now sees a change in tide and see the potential crack in MS and Citrix relationship which for most part didn't exist before when everything was hunkier dory between MS and Citrix when the first agreed to do joint development.  I believe there was reason Microsoft wanted the joint development...  And I believe in the next 3-5 years, the real spots of MS will start to show in this area of technology and market.
Cancel
That statement was made by me (speculatively, I might add), not Microsoft, and was intended to point out the fact that Microsoft now owns a product that will directly compete with Tarpon, when it is released.  Who knows how this may (or may not) affect the relationship between Citrix and Microsoft.  The thing to remember is that we (the general community) really have no idea what is going on inside Microsoft, Citrix or any other company out there.  However, it is interesting to spark discussions on how events like this can shape this industry's future. 

If you ask me, the Softricity acquisition was done to facilitate a faster adoption of Vista when it is released, and to close the competitive gap between Microsoft and Altiris (with their SVS product) - not to "begin the demise of Citrix".

Cancel
Ron,
 
I see lots of people arguing that CPS is better than VDI or that technology xyz is better than VDI etc etc. Correct me if I am wrong but I would speculate Citrix has some 10/15% of market penetration (Softricity is close to 0% so it's not even worth noticing here). I really think that what VDI is all about is the remaining 85/90% of standard PC deployments. Of course there will be some overlaps meaning that administrators that are part of that 10/15% might want to look at VDI (especially with the right tools in place down the road) but I think that it's crystal clear for the majority of consultants that IF THE CITRIX MODEL CAN BE IMPLEMENTED..... that is the best model.
 
So rather than pursuing the discussion of "for that 10/15% of users Citrix is better than VDI" I suggest we should all move to a more interesting "why are there still 85/90%" of standard PC deployments and how can we address that issue?". Not only that: if you consider that (I think) the majority of CPS deployments are done to distribute applications and not the whole "desktop" .... VDI + CPS might become a viable alternative.
 
Now the point is, as I said, why 90% of the people are still using standard PC's? Of course there will be users that will never be ported to a CPS/VDI platform such as very power and mobile users or other things like that ....... but it would be interesting to understand how many of that 90% of users that are still working on a PC .... are doing that because they DO NEED a PC or because they COULDN'T use Citrix ......... which is VERY different as you might depict.
 
Massimo.
Cancel

-ADS BY GOOGLE

SearchVirtualDesktop

SearchEnterpriseDesktop

SearchServerVirtualization

SearchVMware

Close