On Sun, Feb 11, 2007 at 07:27:29PM -0700, Michael Fuhr wrote:
> If using the index would be slower than a sequential scan then you
> don't want the query to use the index. Let's see how accurate the
> row count estimates are and whether using an index really would be
> faster -- please post the output of the following queries:
Also, what's your hardware configuration and what non-default
settings do you have in postgresql.conf? In particular, how much
memory do you have and what values do you have for shared_buffers,
effective_cache_size, work_mem, and random_page_cost?
--
Michael Fuhr
