On Tue, Oct 12, 2010 at 1:07 PM, Tom Lane <tgl@sss.pgh.pa.us> wrote:
> Anyway, if anyone is hot to make COUNT(*) faster, that's where to look.
> I don't think any of the previous discussion in this thread is on-point
> at all, except for the parts where people suggested avoiding it.
I kind of hope that index-only scans help with this, too. If you have
a wide table and a narrow (but not partial) index, and if the
visibility map bits are mostly set, it ought to be cheaper to read the
index than the table - certainly in the case where any disk I/O is
involved, and maybe even if it isn't.
--
Robert Haas
EnterpriseDB: http://www.enterprisedb.com
The Enterprise PostgreSQL Company