> > testinsert contains t values between '2009-08-01' and '2009-08-09', and
> > ne_id
> from 1 to 2. But only 800 out of 2 ne_id have to be read; there's no
> need for a table scan!
> > I guess this is a reflection of the poor "correlation" on ne_id; but, as I
> said, I don't really t
Since noone replied to
http://www.mail-archive.com/pgsql-general@postgresql.org/msg133360.html, I
tried another approach:
I can't cluster the whole table every day; it would take too much (as I said,
table as 60M rows, and I have hundreds of them).
Plus, it wouldn't really make much sense: the
