> What sort of dataset are you indexing that has trillion entries?
It doesn't say
> What sort of dataset has a trillion entries?