Whoops - yeah, we're working on that problem right now - at the moment it's back up.
Was that a O(n!) query killing the DB? :)
It looks like it should stay up now - the traffic from this is much more than we usually get, but it's doing fine now that the server has more resources available to it.
You sure there couldn't have been an algorithm you could have optimized instead of just throwing more resources at the problem? Your solution seems awfully real world.