My datacenter fantasy
As I was driving through the small town in which our district is based yesterday, I was pondering a conversation I'd had with a local fiber company. In middle-of-nowhere Massachusetts, as in much of rural America, broadband is not easy to come by. It's expensive and the kind of really high-bandwidth connections available to urban and suburban customers isn't to be had.
So what's a guy to do? Find partners and share the high cost of serious bandwidth, at the same time leveraging a whole new degree of connectivity. I call this post my datacenter fantasy because the startup capital and interagency cooperation required to make this happen is about as easy to come by as the broadband we're talking about. But I can't help but wonder a few things. With the right partnerships, how much money could be saved long term? How much could we expand access to high-speed Internet? What impact might this have both on education and the local partners?
Let me lay out my fantasy for you...then feel free to tell me where such a thing has worked for you (and I know it's worked somewhere) and if you think the plan is worth pursuing. If it's not, what are the alternatives?
There are quite a few abandoned buildings in our town. The recession took out businesses that had managed to make it through manufacturing downturns. These buildings can be had for cheap money. What if we purchase one, air condition it, upgrade the electrical services, add some server racks, and run several data pipes in from our local cable company (Time-Warner is currently our best bet for high-speed connections with lines up to 15MBps available - this number, by the way, should say something about the state of broadband here, but at least we have cable which is more than can be said for some towns in our area).
I'm not talking about a state-of-the-art datacenter here. We can add xenon fire suppression and tap the local hydroelectric plant later (my tongue is firmly planted in my cheek here, but there would certainly be room for growth and enhancement in the years to come). I'm talking about a location for bandwidth aggregation, a central spot to start several fiber runs, and a building far more suited to handling servers and communication equipment than any other building in a 25-mile radius (the bar, unfortunately, is not set very high on that one).
That's part 1 of the big capital expenditures. Part 2 is running fiber from the datacenter to the town hall, the library, the police and fire departments, the schools, any businesses that want a piece of the action, and the neighboring town with which we are a regional school district (again, to serve EMS, town services, a library, and the schools there).
That's a lot of partners, by the way. Perhaps the most important would be local businesses who no longer want to handle their own servers or are tired of sub-par bandwidth at high prices.
Now comes the easy part. All of those servers that are inevitably sitting in closets, under desks, and stashed in all sorts of places where servers shouldn't live get moved to the datacenter. Ultimately, participation in this collaborative effort would buy access to shared server resources, virtualization, thin clients, and whatever else you can support with powerful servers and lots of bandwidth. For now, the fantasy datacenter simply serves as a clean, secure, high-bandwidth hub for the municipal and business IT needs.
It wouldn't be hard to make some of this bandwidth publicly available either. A central downtown location and some decent WiFi equipment would do the trick. A public computer lab? Mighty easy in a town thirsting for retraining and educational opportunities when unemployment is well above the state average.
Now here's the best part. It's run by students in a new technology education program as a real-world living lab. Students would leave the high school-based program with a full understanding of modern networking, project management, server operating systems and support, and business IT.
Pretty cool, huh? Talk back below and tell me what you think.