Re:Analysis of Google Caffeine.
by bensummers on Wednesday August 12, 2009 @01:34AM
(#757368)
Wouldn't the cluster have a certain minimum size to hold the index data? So it's not going to start at one server and scale up, it's going to start at hundreds.
But of course, I could be completely wrong, not knowing anything about how Google works.
But of course, I could be completely wrong, not knowing anything about how Google works.