Could we make VDI any MORE complicated, please?
Sheesh…can we make VDI any more complicated to deploy? And we wonder why so many organizations just bite the bullet and deploy fully-provisioned desktops? We come up with all of these really cool solutions like Virsto, or Fast Cache, or View Storage Accelerator…or Thin App…or Persona Management…and all we’ve done is complicated the desktop to the point where the “costs” far outweigh the “benefits”.
This is something that I’ve wrestled with for a long time. VDI, in its simplest form, is putting a windows desktop onto a hypervisor, and then putting a “broker” of some sort in front of it to make connecting to the desktop easier and to add some protocol enhancements to make it run better over the WAN. However, that “form” of VDI, while it exists is the wild, has always been an endangered species. It’s just too expensive to deploy VDI like this at scale unless your organization has some hugely compelling reason to overlook the cost of the solution in favor of the benefits.
So, for the last 6 years, the ecosystem has been telling customers (and each other) that deploying Linked Clones (Composer, PVS/MCS, Array-Based) was the way to go to reduce the high cost of the fully-provisioned Virtual Desktops. At the same time, we also have been espousing how much money we were going to save on management costs by going to a single-disk image. Right.
Once you architect, design, and pay for all of the 3rd party solutions that you need to make linked clones work, most orgs have only touched a fraction of their overall desktop use cases. This means that I STILL have to manage the old desktops, and now, have a whole bunch of new desktops that need to be managed a completely different way. New support and operational structures have to be adopted as well.
Did we lower the cost of anything in that scenario? Maybe at a huge scale in some really specific use cases…like call centers, or health care…but tell me that a lot of those use cases shouldn’t have been XenApp/Presentation Server to begin with.
I like VDI. I’m writing this email on my VDI desktop. Doing my demo’s on my VDI desktop. My Fully Persistent Virtual Desktop, that is.
Yes…it IS all about the Apps…and when we lose sight of that, we get mired down in how to make the OS better…when we should be spending our time and energy on better ways to deliver the apps and the functionality that make the world go round.
Now…after wrestling with this over the last several years, I finally think I see a solution to a lot of this insanity in the form of XtremIO (Coming Soon). I’m seeing a hardware solution that is going to allow our desktop architects/engineers the ability to un-architect much of the complexity out of the vdi space. This complexity that has driven up cost, complexity, and risk. This same complexity that has kept VDI from ever truly taking off.
I like XtremIO because it’s simple, and lets us manage desktops a lot more like the physical ones that we do today. It let’s us manage our hypervisors a lot more like we do today. With XtremIO, I feel like VDI has a chance. Without it…we still have great solutions…but they hurt a lot more. Stay Tuned for some really cool stuff…