As many of you will know, the running joke in our industry for the past, well, eight years almost has been that this year, no doubt, without question, slam dunk - is going to be the YEAR OF VDI. I have said it before, and I will say it again – don’t forget what problem(s) you are trying to solve. VDI can sometimes be a “very distracting idea”. Don’t get me wrong, it’s a great idea – but it’s not the whole idea.
To be fair, many of us mellowed our predictions, adding in tap-out phrases like “The Year of DaaS”, “The Year of the Desktop”, and various other niche technologies and concepts to cover our bases. It never happened. Having witnessed over 7 million desktops assessed, I am here to tell you that VDI (in its current technical potentiality) is good for perhaps 20 per cent of the desktops we assess (latency, apps, use case, IO, security, image management, etc).
Let’s do some rough maths - of the 800 million enterprise desktops out there running Windows - 160 million (20 per cent) should be a good fit for VDI. This is based on analysis of a 7 million seat sample size over seven years. Of those 160 million that should/could benefit from delivering Windows via VDI - only approximately 12 million are actually doing so, daily, in production.
Big market - or big problem?
My answer? Both. And to reveal my punch line a bit early in this story - this will be the year of application layering. However, I submit to you that as long as IBM, HP, Dell, Compaq, and Acer have been making machines – it’s been the year of the “Desktop”. That model just solved too many problems to not be the primary content consumption and creation device standard. But we got spoiled. We gave everyone a PC. So, why will this be the year of application layering? Because managing Windows at scale is heavy stuff.
You see my conclusion is that previously we were spoiled. Moore’s law, a computer on every desk, the near instant proliferation (not speed) of 10bt networks, falling capital costs all meant that the running normal for the past 20 years was to configure ‘fat’ machines (and why not, they are cheap!), install every application, program and element a user might use because while we had cheap and fast machines, we had no agility. It was far easier to put everything into that BLOB of a PC versus finding dynamic ways to get there when needed. And so we tolerated this, embraced it even, and every now and then found a shim or a widget to make life a bit easier. But fundamentally scale got away from us. Our new mission is to get scale to work FOR us.
VDI was, and is, an effort to make managing Windows at scale easier (more secure, faster, more leveraged, accessible, etc). You all remember the persistent vs. non-persistent debates where we saw even the most ardent proponents caving in on their previous positions. Why? Because the technology has evolved and now allows us to access architectures we could not dream of a short time ago. However, most of these advancements were in the plumbing of VDI, not the management. This is what application layering solves, we now have a complete set of assets to manage the lifecycle of a workspace, not just its plumbing.
I think most of us would agree that radical innovation does not happen overnight - but rather through sustained and diligent, gentle steps. Storage, IO, user environment management, protocols, hyper-converged infrastructure, security, proactive monitoring, etc. all made huge leaps forward over the past seven years. And gradually, or I would say, FINALLY, application layering is allowing us to attack managing Windows apps at scale. And here is the high-order bit in all of this - application layering should apply to each and every desktop or workspace in your organisation, not just your VDI deployment. One model for application lifecycle management, entitlement and security/compliance.
Legacy application delivery and application lifecycle management only gets us so far. It envisions a static world, managing physical devices with OS’s, applications, data, settings, policies, profiles (stuff) statically installed and bound (very stateful) to the device. We don’t live in that world anymore. Application publishing and virtualisation and cloud delivered apps have forever cracked open Pandora’s box - giving us a glimpse of the power of removing apps from images. Application layering is the single most powerful tool for over 800 million Windows workspaces because of the options users will have to get Windows at scale to begin to work FOR them vs. AGAINST them. The ability to mix and match dynamic single application layers in any way, combination, posture and deliver them in contextually aware or “smart” deployments is, or should be, the focus this year.
When you now look at the options available for application delivery to workspaces of your choosing – be it presentation, installed in image, cloud, virtualised (in the bubble), streamed, or layering – I think you will find that now there truly is no application left behind and, that together with legacy models, or individually, application layering is the final mile in 2016.
It took the best part of a decade but the abstraction of the desktop is now complete. User data, user settings, admin policies, and now, fully, applications have – and can be – removed from the desktop leaving just an OS and the “potential” for a workspace to be contextually delivered on demand.
J. Tyler Rohrer, Co-Founder of Liquidware Labs
Image Credit: Shutterstock/Ollyy