Google's announcement that it has successfully processed its one trillionth link might be linked to the launch of Cuil, a promising search engine built by ex-Googlers.
What does that mean? Google has actually build a database of one trillion URLs but since some webpages actually have more than one URLs (e.g. content being displayed on one page or on multiple pages, syndicated pages, RSS feeds etc), the actual pages on the web is much lower than this figure.
What this doesn't mean is that Google has indexed one trillion pages as some have written; as they put it themselves on Google's official blog, "We don't know; we don't have time to look at them all! :-) Strictly speaking, the number of pages out there is infinite".
As Techcrunch points out, indexing is expensive, very expensive and quite time consuming as well, especially as a significant amount of websites tend to change their content regularly.
As it stands, Google is currently lagging behind newcomer Cuil which has clocked more than 121 billion indexed webpages and Google has acknowledged that the entire webosphere is growing at a similar rate (120 billion pages) every month.
Google took two years to grow from 26 million indexed pages to one billion indexed pages but nearly eight years to reach the trillionth mark.