So lotta chatter out there ’bout so-called “cloud computing” lately. The metaphor is nice and all, since it makes uploading your data to a web-based service sound so etheral. Disembodied and harmless. Plus clouds are nice, fluffy, pretty little things aren’t they.
But in fact it’s not anything like that at all, is it? When you use gmail or google docs or a wiki or any number of web services, your data is being stored in places like this:
(Google data centre in the Netherlands, by Erwin Boogert on Flickr)
And think about just good ‘ol “surfing the net.” Let’s say you are reading the news online — you’re accessing servers in places like this:
(Reuters Data Centre in London England, by .Martin. on Flickr)
And here’s a representative-looking interior shot.
(Cages and Cooling Units, by Webg33k on Flickr.)
What’s that stat about how many servers jump into action when you hit the “search” button on a Google page? It’s close to a thousand machines.
We’re talking about industrial strength computing here. These data centers look like factories, big buildings full of the machines that make the Web. And I gather that inside these data centres the heat and the noise is typically a problem. Again like factories. And they are electricity hogs with a growing appetite for power.
But then again “cloud computing” has a much nicer ring than “factory computing” or “industrial computing” doesn’t it?