Synergy 2015 Keynote – Citrix Cloud Workspace

So today we got the Synergy 2015 keynote, delivered by the usual suspect, Mark Templeton. Here I am at 3:26am in the morning after a couple beers and scotch I had a couple hours ago with old time pals like Shawn Bass, Harry Labana, Ron Oglesby and Michel Roth. Great night and tons of interesting discussions.

After all that I decided to give you my thoughts on what I think is the most important announcement from today’s keynote, the Citrix Cloud Workspace (CCW). Before I go any further let me tell you upfront that my view has nothing to do with the cloud at this stage and that I truly believe the way to go right now is to actually forget about the cloud for a moment. And you will understand why.

For that to happen we must take a look at what Microsoft announced a couple weeks ago: the Azure pack. If you do not know what that is, in a nutshell, it is on-premises Azure. You bring it in-house and now you use the exact same tools/procedures to create your infrastructure on-premises as it would be done off-premises, in the cloud. Why is that important?

First, it eliminates the distinction between on-premises and off-premises as you do things the exact same way, no matter where your stuff is running. That means if one day for whatever reason you decide the cloud is the way to go you already know everything on how to get there. You just point whatever automation/procedures you have to a different container, the cloud. Done.

Secondly, it gives a taste of the cloud to anyone that at this stage will not go for it. It plants the seed on everyone’s head about Azure, showing on-premises how the whole thing works. Without going off-premises.

Finally, everything around the cloud revolves around heavy automation. Many things happen in background, with IT not even really aware of what is going on to get you where you need to be. What I mean by that is, as an example, when I deploy a new application on Azure RemoteApp (ARA) with my custom image (my 2012R2 box with the apps I want to be available), I do not see exactly how Microsoft is actually doing it on Azure. Not that I care. As long as I end up with my apps there and assigned to the users that need them, I am good. The end result is indeed a huge simplification on how we build infrastructure and also how quick we do it. Night and day difference, not to mention the great reduction on human errors as the whole procedure is automated and you barely see it.

Take what I just said and look at Citrix Cloud Workspace. As of today I can have all the backend stuff up there and point all to my XenApp/XenDesktop on-premises. But that is not where the value is IMHO.

If Citrix brings CCW on-premises, I can now deploy my whole XenApp/XenDesktop environment in a heavily automated way. In a couple clicks I can have a PoC up and running for 500 people. The simplification between front end components (RDS Session Hosts, VMs with the VDA, etc) and the backend (StoreFront, DDCs, Databases, Domains) is huge thanks to the connector architecture in use. And again, this erases the line between on-premises and off-premises. If one day you decide you should burst to the cloud or move all up there, everything you have done on-premises is EXACTLY the same thing you would do in the cloud. No more distinction.

This is where Citrix has to go with CWC IMHO. Make the product completely location agnostic, working the exact same way and with the exact same connector no matter if on-premises or off-premises. This will greatly help with multi-domain authentication, SQL connectivity and so on.

This is the way to go moving forward with any solution actually, Citrix or not.

I hope VMware is reading this.

CR

2,919 total views, no views today

Citrix vs. Cloud Platforms. Yawn.

Ok after reading Gabe’s article and then Brian’s take on it, instead of replying I decided to write a whole post about it. That is why you are reading this.

First of all I want to resume Brian’s post for you. I think he should start working for Gartner as he is becoming the master of failed predictions (perfect fit if you want to work for Gartner – not sure if you know this but Gartner has a lot of mediums and Gypsies on staff and is responsible for buying 83% of all crystal balls made in America) and his latest post kind of falls into the same category.

The main idea on both posts is if VMware or another player releases seamless windows apps in their cloud offerings Citrix is fucked.

Here is the deal why IMHO that is not the case and even Brian seems to contradict himself on the post he wrote.

1. The cloud. Oh the cloud. Amazes me to see most CIOs seem to have learned nothing from the whole Snowden/NSA episode. If all corporate systems and intellectual property now lives in the cloud, you just made NSA much happier. The same way Snowden put up their arse, gathering all that information and sharing with the public, don’t you think it would be possible for a Snowden Jr, to get confidential corporate data and give the finger to the NSA and go living in China or Russia with all that info ready to be sold overseas? Do we really think a pharmaceutical company with crazy drugs being developed will consider doing anything in the cloud? Or Lockheed Martin, Bombardier, making Area 51 flying shit , etc? The list of corporations in the Fortune 500 that would be MASSIVELY affected by something like this happening is simply huge. So going to the cloud just makes NSA life easier. Bring the cloud onsite and at least you have a little bit more control and chances to guarantee NSA is kept out of the door.

2. Ok I mentioned bringing the cloud onsite and Brian does mention that, meaning a common platform is there for on-premise and off-premise deployments. But on the same article he also states “Microsoft has started talking about how future versions of Windows Server will be more like “mini on-premises instances of Azure.””. That means this does NOT exist today and only Jesus knows exactly when it will see the light of the day (Nadella or Nutella as I prefer, does not know the answer for that, trust me). So as of today and for at least 3-5 years this is not happening mainstream. Also keep in mind if Windows Server 2016 does have all this shit built-in and working 100% (what is never the case with anything Microsoft releases – for God’s sake they cannot even get RDS to work 100%) companies will still have to go through the exercise of testing and validating such platform what in itself takes years for many Fortune 500 companies. These guys cannot simply change platforms overnight. The FDA would shutdown ANY pharmaceutical attempting to do that overnight. Simple as that. So the reality here is this is still YEARS away.

3. Given point #2, that means a solution, to be called a SOLUTION, and not a HAE (Half Ass Effort) has to support BOTH on-premises and off-premises TODAY. So if someone (i.e. VMware) releases something that only works off-premises, in a cloud platform, we have a problem. What do I do with my on-premises stuff? Ignore it? Choose another vendor to deal with the on-premises scenario only? That is a fucking nightmare. Now dealing with two products and two vendors so I can address my on/off-premises needs. Keep in mind this would still be the case if someone releases a platform that can indeed deal with both scenarios flawlessly within the next year. Why? Because you will still need to test and validate such platform BEFORE going full production with it (point #2). Simple. Common sense here people.

Resuming: as of today and for at least the next two to three years things will still look very similar to what they are today and if you do want to be a leader down the road you must have a platform that deals with the IT landscape of TODAY and with the IT landscape of TOMORROW. Sorry to say but VMware is nowhere near it, in terms of addressing SBC/VDI on-premises and off-premises.

Now if you do not need to test or validate anything, do believe ‘cloudfying’ your whole IT infrastructure is a great idea, and the NSA does not exist, Brian is indeed into something with his article.

CR

26,191 total views, 1 views today

Welcome to the world of tomorrow.

Here I am at BriForum 2009, sitting in a conference room watching Brian talking about Client Hypervisors and thinking about a discussion I had last night with Harry Labana, Citrix CTO for XenApp.

What the future of our desktops, laptops and computing devices will look like and why I think client hypervisors and VDI are the way to go and how they will end up merging at the end.

The thing is, today there is a lot that can be done with the current technology available but as VDI and Client Hypervisors are seen as two completely different beasts the integration I am talking about is not there. Or is it?

The future I see is simple actually. Master disk images for my work machine, home machine, porn loaded machine and so on will exist somewhere, whatever they call that in the future. The cloud is what they are calling it today. It may change to heaven in the future. Or Hell, depending how you look at it and how these guys implement it.

So if I am at the office, my device (a laptop like one) canĀ  connect to my work machine running on a cluster (so here I connect to it using some remote display protocol) – I would do that for example from a location where power is an issue so I use this low power mode on the device to connect (here an actual small app running on top of the hardware built-in hypervisor just to do the remote display protocol part) or if I am on a location I can actually use all the power (and have bandwidth) my device, with a locally cached copy of the image I want to run, just downloads the differences from what is local to what is in the cloud and I run that image locally, at full power and with full access to the hardware (yes, Intel/others will indeed change the PC architecture as we know it today to provide access to GPUs, etc).

When I get home the same takes place. I load my home PC image (just differences) and run it locally. Or again connect to it running somewhere.

Assuming that we get to this point (what I do believe it will happen but not by next year), will there be a market for the traditional TS model we know as of today? I am not sure if it will even be needed.

Hardware architecture changes will indeed allow for much higher densities in the future. Same will happen for power. I am sure Citrix will have their own version of the greatest device of all time that we all know: Mr. Fusion from Back to the Future. That alone will be able to power a shitload of hosted things – remember 1.21 Gigawatts is a lot – probably with minimal heat.

Of course all the management layers to handle all these images running, the applications and patches on them and so on will need to be there. But I guess we will have that sorted out by the time this reality becomes… reality.

Thanks to Microsoft we learned a lot so far on how NOT to design profiles, application deployment tools and so on and if all these companies now working to create this ideal world of the future are watching this and learning the future is bright.

Very bright.

CR

1,068 total views, no views today