We’ve written a lot here regarding XenDesktop’s two provisioning methods: Provisioning Services and Machine Creation Services. Earlier this week, at the Citrix Synergy Conference in San Francisco, there was a session specifically devoted to discussing those two provisioning methods, providing a high level overview of how they worked, the best practices for deploying each of them, and even some guidelines for how to determine which approach is best for your organization. For the benefit of those who couldn’t make it to Synergy - or those who did make it, but would like a better way to share that information with others in their organizations - that session was recorded and is available on Citrix TV. You can view it below:
Category Archives: Citrix Synergy
It’s Been a Cloud-y Week
No, I’m not talking about the weather here in San Francisco – that’s actually been pretty good. It’s just that everywhere you look here at the Citrix Summit / Synergy conference, the talk is all about clouds - public clouds, private clouds, even personal clouds, which, according to Mark Templeton’s keynote on Wednesday, refers to all your personal stuff:
- My Devices – of which we have an increasing number
- My Preferences – which we want to be persistent across all of our devices
- My Data – which we want to get to from wherever we happen to be
- My Life – which increasingly overlaps with…
- My work – which I want to use My Devices to perform, and which I want to reflect My Preferences, and which produces Work Data that is often all jumbled up with My Data (and that can open up a whole new world of problems, from security of business-proprietary information to regulatory compliance).
These five things overlap in very fluid and complex ways, and although I’ve never heard them referred to as a “personal cloud” before, we do need to think about all of them and all of the ways they interact with each other. So if creating yet another cloud definition helps us do that, I guess I’m OK with that, as long as nobody asks me to build one.
But lest I be accused of inconsistency, let me quickly recap the cloud concerns that I shared in a post about a month ago, hard on the heels of the big Amazon EC2 outage:
- We have to be clear in our definition of terms. If “cloud” can simply mean anything you want it to mean, then it means nothing.
- I’m worried that too many people are running to embrace the public cloud computing model while not doing enough due diligence first:
- What, exactly, does your cloud provider’s SLA say?
- What is their track record in living up to it?
- How well will they communicate with you if problems crop up?
- How are you insuring that your data is protected in the event that the unthinkable happens, there’s a cloud outage, and you can’t get to it?
- What is your business continuity plan in the event of a cloud outage? Have you planned ahead and designed resiliency into the way you use the cloud?
- Never forget that, no matter what they tell you, nobody cares as much about your stuff as you do. It’s your stuff. It’s your responsibility to take care of it. You can’t just throw it into the cloud and never think about it again.
Having said that, and in an attempt to adhere to point #1 above, I will henceforth stick to the definitions of cloud computing set forth in the draft document (#800-145) released by the National Institute of Standards and Technology in January of this year, and I promise to tell you if and when I deviate from those definitions. The following are the essential characteristics of cloud computing as defined in that draft document:
- On-demand self-service. A consumer can unilaterally provision computing capabilities, such as server time and network storage, as needed automatically without requiring human interaction with each service’s provider.
- Broad network access. Capabilities are available over the network and accessed through standard mechanisms that promote use by heterogeneous thin or thick client platforms (e.g., mobile phones, laptops, and PDAs).
- Resource pooling. The provider’s computing resources are pooled to serve multiple consumers using a multi-tenant model, with different physical and virtual resources dynamically assigned and reassigned according to consumer demand. There is a sense of location independence in that the customer generally has no control or knowledge over the exact location of the provided resources but may be able to specify location at a higher level of abstraction (e.g., country, state, or datacenter). Examples of resources include storage, processing, memory, network bandwidth, and virtual machines.
- Rapid elasticity. Capabilities can be rapidly and elastically provisioned, in some cases automatically, to quickly scale out, and rapidly released to quickly scale in. To the consumer, the capabilities available for provisioning often appear to be unlimited and can be purchased in any quantity at any time.
- Measured Service. Cloud systems automatically control and optimize resource use by leveraging a metering capability at some level of abstraction appropriate to the type of service (e.g., storage, processing, bandwidth, and active user accounts). Resource usage can be monitored, controlled, and reported, providing transparency for both the provider and consumer of the utilized service.
If you’ll read through those points a couple of times and give it a moment’s thought, a couple of things should become obvious.
First, most of the chunks of infrastructure that are being called “private clouds” aren’t – at least by the definition above. Standing up a XenApp or XenDesktop infrastructure, or even a mixed environment of both, does not mean that you have a private cloud, even if you access it from the Internet. Virtualizing a majority, or even all, of your servers doesn’t mean you have a private cloud.
Second, very few Small & Medium Enterprises can actually justify the investment required to build a true private cloud as defined above, although some of the technologies that are used to build public and private clouds (such as virtualization, support for broad network access, and some level of user self-service provisioning) will certainly trickle down into SME data centers. Instead, some will find that it makes sense to move some services into public clouds, or to leverage public clouds to scale out or scale in to address their elasticity needs. And some will decide that they simply don’t want to be in the IT infrastructure business anymore, and move all of their computing into a public cloud. And that’s not a bad thing, as long as they pay attention to my point #2 above. If that’s the way you feel, we want to help you do it safely, and in a way that meets your business needs. That’s one reason why I’ve been here all week.
So stay tuned, because we’ll definitely be writing more about the things we’ve learned here, and how you can apply them to make your business better.
First Look at Citrix Access Gateway 5.0
At the recent Synergy Berlin conference, Citrix announced Access Gateway 5.0. We have confirmed that, as of now, 5.0 is available for download from the Citrix download site - both as an update for the CAG 2010 hardware appliance, and in Access Gateway VPX (virtual appliance) format. (Note: you will need a “mycitrix” account to download the software.)
One of the things I really like about 5.0 is that it now supports running two 2010 appliances in an active/passive HA configuration with automatic failover. This was a serious shortcoming of the original CAG appliance.
In earlier versions, if you were using the Access Gateway as a general-purpose SSL VPN, you could configure HA of a sort within the Access Gateway client plug-in, by defining primary and secondary Access Gateways for the client to connect to. However, if you were simply running the Access Gateway in “CSG replacement” mode to connect to a XenApp farm without requiring your users to first establish an SSL/VPN connection, you had no ability to provide automatic failover unless you had some kind of network load balancing device in front of multiple Access Gateway appliances. That meant, of course, that to avoid having the load balancing device become a single point of failure, you had to have some kind of HA functionality there as well. By the time you were done, the price tag had climbed to a level that just didn’t make sense for some smaller deployments.
NOTE: This specifically applies to the 2010 appliance. The CAG Enterprise models, because they are built on the NetScaler hardware platform, have always supported operation as HA pairs with automatic failover. Of course, a CAG MPX 5500 also carries a $9,000 list price, compared to $3,500 for a CAG 2010.
Now, with the release of 5.0, you can purchase two 2010 appliances (which will cost you less than a single MPX 5500), and run them as an active/passive HA pair. Thank you very much, Citrix CAG team!
Here are a couple of videos from Citrix TV. The first deals with how to upgrade an existing CAG 2010 to the 5.0 software using a USB flash drive, and then set up the basic system parameters:
The second video shows how to configure a pair of appliances for active/passive failover:
You can access several other “how-to” videos by going to http://www.citrix.com/tv, and searching on “Access Gateway 5.0.”
Citrix Announces XenDesktop 5
Earlier today, at Citrix Synergy in Berlin, Citrix announced XenDesktop 5, which is scheduled for availability in December, 2010. Naturally, we went looking for the “what’s new” list. You can find that list on the Citrix Web site, but, just to save you a few clicks, here’s our take on it.
Most of the user-facing features are evolutionary, as opposed to revolutionary. There have been incremental improvements in devices supported by the Citrix Receiver, the performance of Citrix HDX, user self-service provisioning, and single sign-on. There is also support for XenClient and XenVault, which were recently made available for download as part of XenDesktop 4, Feature Pack 2. But the truly revolutionary, knock-your-socks-off features are on the management side.
Installation and deployment of a large XenDesktop environment is now a snap using the new Desktop Studio tool. Since a video is worth a thousand words, check out the following video demo of Desktop Studio:
But wait! That’s not all! There’s something here for the help desk staff as well, and this may be the coolest part of all. Take a look at a demo of the new Desktop Director tool:
One of Citrix’s stated goals with XenDesktop 5 is to take VDI from “wow” to “how” - to show you how to easily install, scale, and manage a desktop virtualization deployment. Desktop Studio and Desktop Director are huge steps in that direction.
Five Cool Products from Synergy 2010
As many readers know, I spent last week attending back-to-back Citrix conferences in San Francisco. Monday and Tuesday (“Summit”) was for Citrix Partners, Wednesday through Friday (“Synergy”) was for the larger user community. In the coming days, I expect to be writing a lot about stuff I learned there – to the extent that I can without violating the Non-Disclosure Agreement that all attendees agree to as part of the registration process.
Today’s post is about five cool products that I think are worthy of further investigation. I should stress that, aside from Wyse, we do not currently sell any of these vendors’ products, and we may or may not partner with them in the future. So this should not be interpreted as an endorsement other than to say that these products intrigued me and I believe them to be worth looking into.
Wyse XenithTM “Zero Client”
Finally, a non-Windows-based thin-client device with HDX MediaStream video support! I can hardly wait for us to get our hands on one of these for testing. Up until now, if you wanted high performance video, you needed to buy a Windows-embedded thin-client, and install the same Citrix Receiver and plug-ins that you would install on a full-blown desktop PC. And, unfortunately, a Windows-embedded thin-client can easily cost as much as a low-end PC. While I don’t have firm cost numbers yet, I was told it would be “sub-$300” (which I assume to mean $299).
At the Wyse demo, they plugged in the box, turned it on, it auto-discovered the XenDesktop infrastructure and automatically configured itself accordingly, and was ready to use literally in a few seconds. Wow.
Kaviza’s “VDI-In-a-Box”
[Editor's note: Since this post was written, Kaviza was purchased by Citrix, and is now the Citrix "VDI-In-A-Box" product.]
Kaviza has an intriguing product. It won the “Best of Synergy” award in the “Business Efficiency” category. As the product name implies, they make a virtual appliance that handles the provisioning, load-balancing, and management of virtual desktops in a single package. Their original appliance was designed to run on VMware, but the Beta of v3.0 they were showing at Synergy will run on XenServer. They do not require shared storage (i.e., a SAN), or a separate connection broker. When you add more of their appliances, their “grid” automatically reconfigures itself to incorporate the new appliances, replicating desktop template images as required.
They’re positioning this as an SMB solution – up to a couple hundred desktops. If you’re going to grow beyond that, you’re probably going to want the greater storage efficiency of storing your desktop images on a SAN and using the provisioning services of XenDesktop 4. Also, this is specifically a VDI solution, by which I mean a bunch of virtual PCs running on one or more virtualization hosts. As we’ve discussed in other posts, VDI is only one kind of desktop virtualization. If you want the flexibility of being able to leverage all the different kinds of desktop virtualization, XenDesktop gives you that flexibility.
Suggested list price is $125 per concurrent user. Citrix has a VDI-only version of XenDesktop (which does include provisioning services, but does not include any other form of desktop virtualization) which lists for $95 per named user, or $195 per concurrent user. So, taking into account the cost savings from reducing the back-end infrastructure requirements, Kaviza is certainly competitive for smaller deployments, if you’re looking for strictly a VDI solution. Kavisa estimates that, including the virtualization hosts, you’re still under $500/user.
Interestingly enough, Citrix recently made a “strategic investment” in Kaviza, and has licensed their HDX high-performance video technology to them. This suggests that, at some level, Citrix does not necessarily view Kaviza as a competitive threat to XenDesktop 4.
You can view a demo of an earlier version of Kaviza on Brian Madden TV, or go right to the source and sign up for a Webinar on their upcoming v3.0 release.
App-DNA
[Editor's note: Since this post was written, Citrix purchased this product. So they obviously thought it was pretty cool, too!]
Good Lord, if we’d only had a tool like this a few years ago. Several years ago, we worked with a major financial institution that will remain nameless (you know who you are) to build an infrastructure of what was then called Presentation Server that would serve up roughly 300 different applications to roughly 1,000 users. Application Isolation wasn’t available at the time, so we had to do things the hard way. We had a team of several engineers who spent months on application compatibility testing – not only to see which apps would run in a Presentation Server environment, but to see which apps could co-exist in a single server image. It was a huge project, and cost the customer a very large pile of money.
The App-DNA AppTitudeTM software automates the process of application compatibility testing. You give it access to the installation packages of your applications, and it will tell you which Windows desktop and/or server Operating Systems they are compatible with, whether they’re 64-bit compatible, and whether you should be able to package and stream them with XenApp’s app streaming tool or with Microsoft’s App-V. Moreover, if there’s an issue with an application, it tells you what the issue is and makes suggestions as to how you may be able to remediate it!
This product won the “Best in Show” award at Synergy, as well as winning in the “Process Improvement” category. The people I talked to couldn’t give me pricing, but if you’re looking at a major upgrade or migration that involves a lot of applications, this could be a huge time-saver.
Liquidware Labs
Their Stratusphere FitTM product was a Best of Synergy finalist in the “Business Efficiency” category (the category that was won by Kaviza). This is a VDI assessment tool. It will monitor and log a bunch of desktop OS and user performance metrics, looking at network usage, application usage, disk and memory utilization, graphics intensity, disk IOPS, network latency between the current desktop location and the data center you’re hoping to move it to, etc.
After gathering information for a while (a minimum of two weeks is recommended), it will spit out both detail and summary reports that will identify good, fair, and poor candidates for virtualization, identify potential problem areas, and help you size the back-end infrastructure that will be needed to host all of the newly-virtualized desktops.
The cost of a time-limited license (90 days, if memory serves me correctly) is roughly $7 per user. Look at it this way: You can design your VDI hosting environment by the seat of your pants, and probably end up either over- or under-building the infrastructure, or you can spend a little bit of money to develop some hard data to guide the design decisions. If it helps you avoid design mistakes, and helps insure the success of your VDI project, that’s probably money well spent.
Unidesk
The Unidesk product competes directly with the provisioning services component of XenDesktop 4. Why, you may ask, would you want to pay extra for a third party product instead of using the provisioning functionality that comes with all versions of XenDesktop 4? Here are some possible reasons:
- Unidesk integrates patching and version management into their provisioning tool.
- Unidesk can deliver boot-time drivers such as antivirus software, VPN software, and printer drivers as components that are separate from your master OS image.
- Unidesk integrates application management into their provisioning tool, including applications that have been packaged for streaming via XenApp, App-V, or ThinApp.
- The big one: Unidesk treats user-installed applications as part of “user personalization” – yes, you can provision from a single master OS image and still allow users to install their own apps. (And you can also – relatively easily - repair the damage when a user installs an app that breaks something else.)
In some organizations, user acceptance will make or break a desktop virtualization project. In a native XenDesktop 4 deployment, if you want to allow the user to install applications, you have to dedicate an OS image to that user. If this is a requirement for a lot of your users, you’re going to burn up a lot of expensive SAN storage. If internal company politics will allow you to lock down the corporate desktop, great! Your life will be much easier. And, as we’ve observed elsewhere, XenClient promises to address this by giving the user multiple desktops: a corporate desktop that’s locked down, and a personal desktop where they can install their own applications. But if you are forced, for whatever reason, to allow your users to install their own applications on top of the corporate desktop image, Unidesk could save you a bunch of storage space, and maybe even your sanity.
