Comment: Even that, doesn't justify it. By far.

(See in situ)

In reply to comment: Does that $634M include one (see in situ)

Cyril's picture

Even that, doesn't justify it. By far.

Even that, doesn't justify it. By far.

I can't think of any need to buy or build a full data center a la NSA for an application like that, besides deliberately wanting to waste money by the loads (which they may very well have gone for, granted, I could be wrong).

Guys like Rackspace have plants like one in Texas where the cooling fans alone weigh in the two or three hundred metric tons, with thousands of server blades underground to serve dozens (if not hundreds) of thousands websites, with front ends AND backs ends for their customers - from the single amateur to the biggest corporates.

And they're probably not even among the biggest or the most competitive. Yet, I bet they'd be competitive enough to offer really powerful hosting solutions to serve in the trillion new records a year for functioning costs less or much less than $100,000 (a year) - with 24/7 human support and daily backups of the whole shebang, be it with 10 or 100 nodes for the app servers (which I would very grossly guess is the ballpark we're talking about).

And that's from recollections of their capabilities working with them (as customer) in 2009. Their web based server admin dashboard (down to the H/W firewall rules) was already kicking ass, btw - but don't know how is it today.

"Cyril" pronounced "see real". I code stuff.


"To study and not think is a waste. To think and not study is dangerous." -- Confucius