I've got a pretty big database (~30 gigs) and when I do a pg_dump, it ends up only being 2 gigs.
The database consists mainly of one very large table (w/ a few varchar columns) which, according to pg_relation_size() is 10 gigs and pg_total_relation_size() is 26 gigs (we need to drop some indexes there.)
I'm just trying to get my head around the pg_dump being an order of magnitude smaller than the darned database itself. I would thing that the db would offer more efficient encoding for a lot of stuff vs. an ascii file.
Thanks,
Carson