Citrix Licensing – Deciding between concurrent and user/device licenses

Citrix XenApp and XenDesktop are available in two general licensing models:

  1.  Concurrent licensing. This model is intended for one connection to a virtual desktop or unlimited apps for any user and any device –  a license is only consumed for the duration of an active session. If the session disconnects or is terminated, the license will be checked back into the pool.
  2. User/Device licensing. Under this model, the license is either assigned to a unique user or shared device. If assigned to a user, it allows that single user unlimited connections from unlimited devices. If assigned to a device, it allows unlimited users, unlimited connections from that single device.

The User/Device license is typically half of the price of a concurrent license and can be an attractive model for organizations that follow a “traditional” work schedule (as opposed to shift workers in manufacturing or healthcare, where they may be a large number of individuals, but only a fraction of which are concurrently using the XenApp or XenDesktop environment.)

Internally, and this is the topic of this article, if Citrix XenApp / XenDesktop is configured for the user/device license model, the Citrix license server has to decide whether to assign the license to a user OR to a device. These are two different things, although customers purchase a user/device (as in user SLASH device)  license. So, how does this work?

Assume I, florianb, log into my organization’s environment and launch a session. At that time, a user license is consumed. I can run as many sessions from as many XenDesktop sites that share the license server as I like and use as many devices as I care to – it’s still one user license.

Assume that one of the devices I use is a shared thin client in the office. An hour after I leave, my co-worker Alex uses the same client to access his virtual desktop. Citrix internally then marks that particular thin client as a shared device and it consumes a device license. Theoretically, I could have 100 employees each use the same thin client and only consume a single  user/device license.

It becomes apparent that the recognition of shared devices is an automated way for organizations to minimize the number of licenses they need.

Most of us, however, have a mix of environments, so Citrix is calculating the total number of user/device licenses as follows:

# User/Device licenses = (# of total users) + (# of shared devices) – (# of users who only access from a shared device)

Makes sense?

Here’s a simple example:

User/Device Devices Used User License Consumed? Device License Consumed?
Paul Client01 No, because Paul is only using a shared device (Client01, which is also used by Florian, Alex, and Amanda) N/A
Florian Client01
Florian’s PC
Florian’s iPad
Florian’s Laptop
Yes, because he is using one or more non-shared devices N/A
Alex Client01 No, because Alex is only using a shared device N/A
Amanda Amanda’s iPad
Client01
Yes, because Amanda is using a non-shared device (her iPad) N/A
Client01 Used by: Paul, Florian, Alex, and Amanda N/A Yes – because Client01 is used by more than one user
Florian’s PC Used by Florian N/A No – because Florian is consuming a user license so he can use an unlimited number of licenses
Florian’s iPad Used by Florian N/A No – Florian is consuming a user license so he can use an unlimited number of devices
Florian’s Laptop Used by Florian N/A No – Florian is consuming a user license so he can use an unlimited number of devices
Amanda’s iPad Used by Amanda N/A No, Amanda is consuming a user license so she can use an unlimited number of devices

 

So, in this example, we would need a total of 3 user/device licenses, even though we have 4 individual users and 6 individual devices in the mix. Given that the price point for a concurrent license is 2x that of a user/device license, this small sample organization would absolutely benefit from user/device licensing as they may need as many as 4 concurrent users licenses.

The Citrix license optimization definitely works in the customer’s favor and the license allocation happens on a 90 day schedule for user/device licenses (i.e. the license of a user who is no longer in the organization or a device that is no longer in use get automatically released after 90 days or can be released immediately with a license management tool under terms of the Citrix EULA).

However, it can be a little difficult to predict what an organization might need. Lakeside SysTrack is a great tool to look at all sessions (say in an existing XenApp concurrent environment) to determine if a trade-up to user/device licensing would make sense. To illustrate the point, I’ve mocked up a quick and easy dashboard in SysTrack’s dashboard builder to look at one of the many environments we’re running internally.

license_dashboard2

 

In this particular example, our peak user concurrency was 11 and we would have needed 29 user/device licenses. We’re better off staying with concurrent licensing in this example.

Equally, if a traditional desktop environment is being assessed, SysTrack can make the choice between concurrent and user/device licensing very easy.

 

Florian

Twitter: @florianbecker and @lakesidesoft

Email: florian.becker@lakesidesoftware.com

On the web: www.lakesidesoftware.com

References/Notes:

  • While Citrix has reviewed this blog for accuracy at the time of this writing, Lakeside Software cannot make any representations on behalf of Citrix. Please always check with your authorized reseller, Citrix account manager and on citrix.com for the latest updates in product and licensing functionality.

Why Would Anybody Want To Do That (To Me)?

My good friend and former colleague Chris Fleck (@chrisfleck) is a well-known enthusiast for mobile work styles and the latest gadgets. His recent blog post asks his readers why they are not using their iPad or other tablet for all of their work. His embedded poll lists all the right reasons, but the first thought that came to my mind was “Why would anybody want to do that?” There are, of course, iPad enthusiasts who would like to use their device for absolutely everything and show that it’s possible to do so, but I personally don’t think that this is realistic for most of us.

I believe that many of us are most productive in our jobs when we have the best resources available for the job at hand. If I am mostly attending meetings, taking notes, or firing off short emails, I will  be in heaven with just a tablet. In my personal case, I have a pretty powerful workstation with dual monitors, plenty of memory and CPU and fast SSD’s. Most of the time I don’t need all of that power, but it allows me to run complex data queries, run separate VMs with server images, etc. A big factor in my productivity is the full keyboard/mouse and the large screen real estate. So, that’s my ideal rig. But, it’s not mobile. So, the more mobile I need to be, the more capability I will need to forfeit. The following chart, which is neither scientific nor necessarily representative of your environment,  may illustrate the point a little better:

WWADT-01

I think that each task or application falls somewhere on my arbitrary capability scale. For example, writing a blog or creating a presentation would be in the 30-40 range, meaning that a laptop is good enough for the job, but the tasks becomes tedious on a tablet and close to impossible on a smart phone, even if I have the ability to access a virtual desktop or application from it. In comparison, clicking on an approval link sent to you in an email by your ERP system requires not much and the smart phone is the preferred way of handling that task for many of us today.

Why do we then use these mobile devices if the productivity can drop so much? It’s obvious: Because the alternative is that we wouldn’t be able to do anything at all while on the road, on an airplane, in the car, bus stop, at a bar, etc. There clearly is a factor of “good enough” here that applies and people accept inferior user experiences if they are offset by another advantage such as mobility. A great book by Clayton Christensen describes this in great detail and with many examples unrelated to the IT industry. (http://www.claytonchristensen.com/books/the-innovators-dilemma/)

A different way to look at the same problem is to determine how much capability I may need on average and what my peak demand may be and then plot that against what I have available. Again, the scaling here is somewhat arbitrary.

WWADT-02

In the example above, I have plenty of extra resources available on the desktop and it appears that this particular example of Thin Clients with VDI provides just about enough of what I need to do. And just in case you wonder: I consider the Thin Client / VDI solution to be somewhat mobile because it allows me to roam around in the office  and in between offices and still connect to the same desktop and applications. But what if it didn’t live up the capability I need? What if my IT department were to provide me with something like this:

WWADT-03

I could do most of my work, but I would have to spend plenty of time waiting for resources on the machine, be limited to only a few concurrent applications, and even simple tasks may require significantly more time and patience than necessary. For some applications, the user experience may be poor enough that they would not be usable at all.  I would probably get frustrated sooner rather than later and seek alternative ways to boost my productivity – maybe by bringing in my own laptop or PC. To paraphrase my initial question, “Why would anybody (IT?) do that to me”? That’s mostly for a number of reasons, but the primary one is that today’s IT departments are flying blind when it comes to End User Computing. It’s not IT’s fault, but IT has very limited insights into what users are actually experiencing. It’s been traditionally very difficult as a practical matter to do comprehensive surveys of the end user community. To make things worse, things are constantly changing and a static IT has little chance to keep up. It would not be right to point fingers at IT though – traditionally, there are very few tools and processes available to determine the end user experience or gather the data that would establish a baseline ahead of any BYO or virtualization initiative.

That’s where SysTrack comes in. It’s the only true Big Data for End User Computing toolset to let IT get real insights into the user experience, design the technology behind transformational initiatives, and stay on top of changing requirements.

The following is just one of many data visualizers that clearly show – in hrs. per week – how much and in what areas the user experience is impacted:

WWADT-04

The system on the bottom of the graph has more than 10 impacted hours per week – mostly due to CPU constraints (blue), while the one right above it experiences mostly disk/IO/storage limitations (red.) You can see all the different limiting factors in the chart legend and SysTrack allows the administrator to drill deeper into the specific system, session, or user and get to the application and process level along with their resource consumption.

This is just one small example how SysTrack can the feeling of flying blind into 100% visibility.

Florian Becker

Florian Becker

Twitter: @florianbecker and @lakesidesoft

Email: florian.becker@lakesidesoftware.com

On the web: www.lakesidesoftware.com