On Fri, Sep 16, 2005 at 04:45:57PM +0200, Peter Eisentraut wrote:
> The planner thinks your query will return 414 rows, so it thinks the
> sequential scan is faster. In reality, your query only retuns 28 rows, so
> you need to create better statistics, either by running ANALYZE or VACUUM (or
> both) or tweaking the statistics parameters of the columns.
I did wonder about the planner stats, so I had run ANALYZE on the
database with no change. I just now ran VACUUM (and VACUUM ANALYZE),
and again see no change. Perhaps my table is just too small for this
test.
--
Bill Moseley
moseley(at)hank(dot)org