For the past few years I’ve had an idea that I’d repeat over and over to anyone that would listen to me. I’ve been given a hard time about it on twitter, have been told about every technical reason why it’s not possible but in the end I stick by this idea: I want micro servers for virtual desktop (VDI) deployments.
My reason for this is simple. When a desktop dies, productivity stops, people get upset, and worst of all work will inevitably be lost. This is a fact of life and thus IT should be designing systems that accept the eventual failure of a single server and they should design the solution so that this failure minimizes user impact. Putting 100-200 VMs on a server may seem attractive from a cost stand point (due to the current way of thinking). But pulling the plug on 100-200 desktops is not something that should be considered acceptable. Instead, I am of the belief that we should be designing servers that have fewer users on them. In my perfect world a server should be completely disposable, costing significantly less than modern servers and host around 10 users (or less).
In today’s world this idea is insane. Servers are expensive and space would be a major concern because a small server burns at least a 1U, some compact servers allow two servers per U. If I wanted to build a system to support 2000 users and only have 10 users per host it would never work from a space standpoint. Okay, fair enough, but that argument just led me to believe we need smaller servers, much much smaller servers, and a little over a year ago I learned about HP’s project Moonshot.
Even though server hardware isn’t a big focus of mine I opted to attend one of HP’s early briefings on this project. I liked the size of the server, it was what I was looking for, but at that time that’s where the honeymoon ended because this server was going to be based on ARM. ARM wouldn’t work in my world of virtual desktops, I need x86 and I need VT, without these my ideas where DOA. Well I should have kept paying attention because this week I learned that HP made some major changes that bring my dream back to a reality. Most importantly I learned that Moonshot is now using an Atom processor that supports VT (or so I’m told). This at least makes my thesis technically possible, however at this point I still have more questions than answers, but I am excited at the possibility of what this brings for desktop virtualization solutions.
I love the idea of spreading out the load of virtual desktops. I don’t know if HP is thinking about virtual desktop, but I think they should if the price is right. My hope is that as this project becomes a reality I’d get to test it and see if it’s possible to deliver a decent Windows experience on this processor (my previous experience with Atom on desktop is anything but a good user experience). Regardless of whether it is possible today, I like this direction of thought and I know that technology has a history of getting faster and cheaper. So who knows, maybe one day soon my crazy nut brained idea might become a reality.
Feel free to follow me on twitter if you wish to continue the conversation @gunnarwb
Disclaimer: This blog is my personal musings on a specific technology. These views do not reflect Gartner as an organization (it takes a lot more work for me to formulate a fully hashed out Gartner opinion than the 10 minutes it took to write a blog). I do not endorse any vendors mentioned in this article, the only thing I endorse is Gartner. Gartner is pretty awesome. I mean they actually pay me to research technology. How cool is that?! Okay, I forgot this was supposed to be a disclaimer. Legal jargon, legal jargon, don’t sue me. The end.
View Free, Relevant Gartner Research
Gartner's research helps you cut through the complexity and deliver the knowledge you need to make the right decisions quickly, and with confidence.Read Free Gartner Research
Comments or opinions expressed on this blog are those of the individual contributors only, and do not necessarily represent the views of Gartner, Inc. or its management. Readers may copy and redistribute blog postings on other blogs, or otherwise for private, non-commercial or journalistic purposes, with attribution to Gartner. This content may not be used for any other purposes in any other formats or media. The content on this blog is provided on an "as-is" basis. Gartner shall not be liable for any damages whatsoever arising out of the content or use of this blog.