October 26th, 2006
Why mobile data centers matter
I was listening to Gillmor Gang last night while I was getting snowed on at my son's football game. Part II (and for the record, I hate these insipid multi-part shows as much as everyone else) of the Black Box Gang discussed Sun's new mobile data center. There were arguments why this was cool and why it didn't matter. Jason Calacanis didn't get it and kept asking why this was important given there's so much data center space sitting idle and blades were making data centers denser.
The answers others gave to Jason's objections missed the mark. There are two important reasons why these mobile data centers matter: latency and net neutrality.
When I was Utah's CIO I realized that latency, not bandwidth, was the largest threat to the kinds of services that government and others might want to roll out in the future. We worked hard, with little success unfortunately, to establish regional peering points in Utah.
Mobile data centers can help overcome this limitation, at least for the big players. If Google wants to deliver video worldwide, putting mobile data centers in strategic locations will vastly improve the quality of service that their customers see.
Net neutrality is, perhaps, a more interesting problem. Mobil data centers serve as a trump card that large content providers can play if they are threatened with increased costs for "enhanced" (i.e. not intentionally damaged) service. I don't think this can counter every threat, but dropping mobile data centers in strategic locations can route around particularly obnoxious networks.
Steve predicted we'll see these things helicoptered onto the top of skyscrapers in the financial district of NY. I don't doubt that's true. But I also think we'll see these things springing up in less obvious places as large content shops use them to solve real network problems.
Phil Windley is an Associate Professor of Computer Science at Brigham Young University. See his full profile and disclosure of his industry affiliations.
