We at F5 – like most collectives of geeks - are constantly discussing the wide array of IT boondoggles that are out there, looking at which ones hold water and which are just passing fads. Often we’re debating which are passing fads. Today I received an email to a small group asking if any of us had tried out the augmented reality stuff out there. I haven’t, but that gives you an idea of the edge that is sometimes taken.
And it is that time of year where every pundit and their uncle is making predictions about what will happen in high-tech next year. It is easy to forget that high-tech has long suffered under the “this year is the year of X!” syndrome when each year we hear it in forecasts and predictions as if it was the gospel truth, even if anyone practicing the craft on a regular basis knows that the prediction is wrong. I can recall years when Disk was Dead (almost every one since the turn of the century, in fact), years when iSCSI would eliminate Fiber Channel (abused so much that when iSCSI + NAS finally does, no one will notice), years when ISDN was going to take over the world (and two decades later, It Still Does Nothing), and predictions that in 2010, hardware load balancers would disappear. Funny, I still work for a rather successful company that makes them.
Ever notice that the people who make these predictions will tell you how right they were for each one that comes true, and will never again mention the ones that didn’t? One year at Network Computing we at least went back and said “here’s where we hit, here’s where we missed last year”, most people don’t even do that.
But I digress. I saw an interesting statistic in a White Paper that Lori wrote that quoted an InformationWeek Analytics survey claiming 77% of you were implementing or about to implement Virtual Desktop Infrastructure (VDI). Really? I disbelieve. The margin of error in that statement when taken in context of the entire marketplace is at least 50%, I’m guessing. Unless a whole lot of you are implementing VDI already and forgot to tell me, which I suppose is possible, but then the sample of people I talk to all the time must be luddites in disguise I suppose – because none of them are. Yet.
For me, the intriguing thing about VDI is that it enables remote access from alternative devices. That’s pretty cool. The payroll manager can field calls about your $0,.02 check while he’s watching Sunday night football, or even on-site at the NASCAR race. Support doesn’t even have to get dressed, and has a standardized and secured method to access systems.
But VDI, no matter your vendor of choice – or whether you’re using application virtualization, desktop virtualization, or whatever, puts a heavy burden on the network. Even the most optimized VDI vendor is still not a networking company. They’re an application company, their application being VDI and their application protocols being IP protocols designed primarily to communicate large volumes of data between virtualized desktop and current host. The impact of the network is large, simply because applications are still making all of the network connections they were in a non-virtualized world – to the databases, AAA servers, etc – while the virtualized desktop is sending drawing information (in one form or another) to the host and receiving keyboard/mouse input back.
With the proper architecture, this and many other VDI concerns - like a public authentication server and massive encryption to send data back and forth over the Internet – can be overcome, it just requires some homework and forethought. Of course I think you’ll be happiest with F5 gear tot optimize and secure your VDI installation, but having seen architectural diagrams and heard back from customers, I can attest to the fact that my sentiment is based upon facts, not marketing BS. Your deployment can gain a lot with little investment from F5, but you can find other solutions to most of the problems our gear solves, so I’ll let you decide what’s best for your environment.
And VDI holds a lot of promise. Allowing access to work from several locations, presenting the same desktop whether the user is in the San Jose or Dublin office, allowing access from portable devices for those who work from a variety of locations, and the ability to control what’s on the VM without having to guarantee that every host is locked down to corporate standards. Of course, there’s the ability to archive or delete desktops too, which isn’t a huge driver, but certainly is attractive. And most of all, the ability to update 100s of desktops without having to touch a single desktop machine. Operations should be drooling, even if you’d have to sell it to users with the promise of easier access and desktop portability.
Am I making a prediction about 2012? Oh heck no, I’m not currently in a job that requires I make predictions, so I’m not. I’m just pointing out that you should consider it, since VDI is rapidly coming of age, and the use case of tablets makes it more appealing to users. This time of year isn’t a time of predictions for me, it’s a time of holiday cheer. And I certainly wish it to all of you over the next month or so.