Brian answers the BriForum "What would you ask Brian?" questions

BriForum 2014 Boston just wrapped up. One of the cool things we did this year was to include the following questions on the registration form: If you had a chance to ask Brian Madden one question at BriForum regarding desktop or application virtualization, enterprise mobility, or consumerization, what would you like to know?

BriForum 2014 Boston just wrapped up. One of the cool things we did this year was to include the following questions on the registration form:

If you had a chance to ask Brian Madden one question at BriForum regarding desktop or application virtualization, enterprise mobility, or consumerization, what would you like to know?

So here's the results of those questions:

All-Flash versus Atlantis + Traditional SAN. Is it really worth it?

Obviously every situation is unique, but if I'm going to give a blanket answer, I'll say this: In 2014, you don't need to go to all flash / all SSD in order to get performance, as there are lots of great software storage products that can give you the performance you need with a combination of spinning disk, RAM caching, some flash, and software smarts.

That said, remember last month I wrote that all storage vendors are really software vendors, so it's hard to make a blanket statement that one type of storage is always better or more worth it than another. There are all-flash vendors with great software that might be able to put together a solution that gives you what you need for less cost than solutions that including spinning disk. But there are also vendors who include spinning disks in their solutions that might give you more bang for your buck.

Really it depends on what you need (in terms of capacity, IOPS, data protection, deduplication, etc.). And if you find a solution that works for you, frankly I don't care whether it's flash, RAM, spinning disk, or a million gerbils with punchcards. If it works for you, go for it!

Since Desktop Virtualization essentially changes the way end users interact with their workspace, it changes to the entire paradigm of end user enironment management. Our current scenario has lot of focus is on desktop virtualization technology and its performance, but I feel there is less focus on the "Operationalization" aspect of Desktop Virtaulization environments. So my question to Brian is "How important is the operationalization aspect for a sucessfull deployment of DV technology?" and "What are good practices to implement DV operationalization in enterprises?"

Way back at BriForum 2006 Washington DC, Ron Oglesby did a session called "Operationalizing your VMware Environment." (A video of that session is here.) I agree that this is something that a lot of us don't focus on and that it would be a great session (or at least article) for the future!

In the meantime, I'm a proponent of the idea that VDI is just a form factor change, and that if you want to be successful with VDI, you need to focus more on the Windows management aspects of it and less on the VDI technologies themselves. (So in that sense I totally agree with the premise of the question.)

That said, we have already operationalized our existing desktop estates for the past twenty years, so one way to answer that question is that we shouldn't try to operationalize VDI itself, rather, we should just keep doing what we're doing. Frankly most of the alleged management "benefits" of VDI are things that we can do with traditional desktops. (Cloud-based syncing for files, app packaging and delivery, single-click restores, user profile virtualization, etc.) So with VDI, I wouldn't focus on trying to make your environment better to manage, rather, I'd just get the technology rolled out and then keep doing what you've been doing all along.

How do you benchmark VDI client performance independent of the server or backend storage so that appropriate zero clients or thin clients can be purchased?

The quick answer would be to set up a monster single user environment on a powerful server with unlimited IOPS and then get a bunch of clients and test them. That way the server, storage, and backend would not be a bottleneck at all. If you do that you can find the most performant thin or zero client out there.

But then I'd have to ask, "Is that realistic?" What are you actually testing? Sure, you'll find the client with the theoretical best performance, but will that relate in any way to what your users will actually be using? I'd be more inclined to test out different clients with your actual backend environment so you get a more realistic picture of what your actual user experience would be like.

I am a good Citrix Consultant, but I want to make the jump to the next level and be a great consultant. Are there any resources that you could recommend to help me make that transition.

Read the Windows 7 Resource Kit cover-to-cover. (1700 pages but worth it.)

Then browse through the 70+ videos from BriForum 2014 London and BriForum 2014 Boston that you have access to as an attendee. There's also lots of good stuff in the 500+ videos we have online from the past 14 BriForums from 2005-2013. (We put all of 2012 and 2013 on YouTube too.)

But seriously, start with the Resource Kit.

There is no question that the technologies are here, for the most part, to let us work the way we want to. How do we move an organization’s IT mindset to ‘user-centric’ IT?

I have no idea. I'm up for ideas though if anyone wants to share them in the comments?

Seriously, my experience so far has been that I have failed to convince anyone who didn't "get it." I talk and talk about FUIT and consumerization and everything, but a lot of companies just view it as an unnecessary expense.

Hell, I've been trying to convince TechTarget to make this transition since joining them six years ago. My team uses Macs and Dropbox and Google Docs and everything, but just about every other of the 700 employees uses a low-end laptop that is super slow. I watch my coworkers lose productivity in little 30-60 second increments dozens of times every day. We have a web-based VPN that is anything but user friendly, and the whole company is based around file shares (that you need the VPN to access). We are the epitome of emailing documents and outdated thinking, but so far they haven't changed. (They think that because employees use unmanaged iPhones now that they're "embracing" consumerization.) So if I can't convince the company I work for to change their mindset, man, good luck!

VMware and AWS walk into a bar, who buys the drinks?


What does VDI outsourcing/hosting mean for the corparate VDI admin's career. I personally left a large corporation in part because I personally believe that the VDI/app delivery diehard has a limited lifespan on the corporate payroll despite the large demand for the skills today.

The good news with VDI is you still have to manage Windows, and Windows desktop apps will be around forever (at least in the enterprise). If I was an IT desktop guy at an enterprise I would push us to DaaS (for the use cases where VDI makes sense) just so I could get out of the VDI management business. That way I could focus on all the new stuff, like figuring out how to make enterprise apps and data usable from all devices, focusing on collaboration, etc. (Of course that requires an enlightened company which is certainly not everywhere.)

My general view though is that the march of technology means that there is a lot of stuff that we used to do that we won't anymore, but that's ok. I'd try to view VDI not as a hack for dead technologies but rather as a way to bring those older technologies into the modern way of working. The whole concept of IT changing to an enabler is really cool and fun.

But again, it really depends on the company and the culture, and some places the VDI admin is a dead end job.

But if you know Windows desktops inside and out, you'll have a long career.

When was the last time you designed, built and deployed a virtual desktop infrastructure for a large scale enterprise beyond ~5,000 virtual desktops.


The consulting I do now is at various phases of projects, but I've never been dedicated to a project start-to-finish from design all the way through deployment. Obviously an environment greater than 5000 users is a multi-year project, so I'd have to disappear to do it.

That said, I'm still involved in a lot of projects of this size from a consulting standpoint, but I'm more popping in here and there to talk about what others are doing in the space rather than being the actual guy leading the project.

Where do you find that the current software tools you use to manage your virtualization solutions fall short?

Two places:

First is integration. There are so many great tools on the market, but I still feel like none of them are really integrated together. So you buy the best user profile manager, but it doesn't really hook into your VDI admin console. You do backups with another tool and security with another tool. Even the best-managed VDI and RDSH environments are all held together with chewing gum and duct tape. (I like what companies like RES Software are doing with their automation stuff as a starting point.)

Second is I still haven't found a monitoring suite that I love. I still feel like a lot of these things are just glorified perfmon log aggregators. (Years ago Kevin Goodman said they were all YAMs, for yet another monitor. I still think that applies today.) To be clear, there are some great ones out there (shoutout to Lakeside Software!), but I still feel like what these things actually do and our fantasy of "tell me when users are about to complain and what I need to do about it" isn't where I thought it would have been by now back in the day. We need some AI or something.

Where is this technology going, which vendor is on the cutting edge, which vendor is stable/reliable?

It's hard to answer this in a single article. We talked a lot about where things are going in the final chapter of our VDI book. The short answer for Windows desktop applications is that (1) they will be here forever, and (2) we will move beyond the full Windows desktop as the "package" that we manage.

For cutting edge, I love that the storage vendors now let us have fully persistent VDI with good performance for a decent price today. If I'm building new VDI environments on premises, I love hyper converged vendors (storage and CPU in the same box) like Nutanix, Pivot3, and Maxta (who were all at BriForum). I like what PowWow is doing in the app refactoring space to make Windows desktop applications more useable from phones and tablets. I like what FSLogix and CloudVolumes are doing to move beyond app virtualization to allow lots of apps to run side-by-side in the same image. Oh, and Moka5. (More on them in the next answer.)

As for stable and reliable, I don't know. None of them? Certainly VMware has been better at this than Citrix recently (which makes it even more exciting that they now have app publishing and RDSH support in Horizon 6.)

Microsoft is still Microsoft.

While it seems obvious that DaaS is cutting a path through the future, it seems unlikely for mass adoption for some time to come. VDI is great, but also comes with quite a large commitment. What are ways that companies are using Application Virtualization (in all its iterations) to deploy solutions that enhance the user experience while centralizing support?

I definitely like the idea of using some type of new way to deploy and manage Windows desktop applications, and I like that all the solutions today also work on traditional desktops and laptops. So in addition to traditional app virtualization like App-V and Symantec, I like FSLogix and CloudVolumes (as I mentioned above), and VMware Horizon Mirage.

I also really like what Moka5 is doing in the client-based VM space. (They're one of the most underrated vendors in our space today. I don't know if people poo-poo them because of the bad connotation of "client hypervisors" or because they had a nutty CEO (my opinion, not libel) for a long time or because they didn't get bought or what, but there stuff is great. It gives you modern management, layering, etc. without VDI. Love it.

Who is leading the virtualization infrastructure war? Are Citrix and Microsoft in bed? Will VMware continue to rule? Or is there someone hidden that will take everything away from everybody?

The virtualization war is now between VMware and Microsoft, (but I guess everyone knows that).

Citrix and Microsoft are most certainly in bed.

VMware will continue to rule for the near future at least.

I can't imagine anyone taking everything away from everybody in the enterprise space anytime soon. (Even VMware took 15 years to get to where they are today.)

Who is your barber? No, but really, what new vendors in application and desktop visualization solutions should organizations be on a lookout for? Anyone out there on the verge of improving HDX, RDP or PCoIP (yes in that order :)?

My barber is Eman Benisano. He's awesome.

As for new vendors, I like PowWow, FSLogix, CloudVolumes. I like the hyper converged storage vendors. I like DaaS. (Seriously.) I'm really excited about the new version of Windows Server and Citrix Workspace Services. And I really like Horizon 6. So easy to use.

I'm not sure about anyone improving the protocols. The people in the position to do it are the telcos really, especially as more stuff moves to the cloud.


So that's it. Thanks for all the questions. We did this at BriForum 2014 London as well. Check out those questions and answers here.

And feel free to post more questions, add more info, or refute these answers in the comments below!

Join the conversation

1 comment

Send me notifications when other members comment.

Please create a username to comment.

Microsoft aren't buying the drinks - they're selling them. And they don't allow two different customers to be served from the same bottle.