pgsql(at)mohawksoft(dot)com writes:
>    ->  Index Scan using rt1_zipr, rt1_zipl on rt1  (cost=0.00..121893.93
> rows=30835 width=302)
>          Index Cond: ((zipr = 2186) OR (zipl = 2186))

>  zipl      |        925 |
>  zipr      |        899 |

Those n_distinct values for zipl and zipr seem aberrant --- too low
compared to the estimated rowcount you're showing.  What are the
true figures?  Also, how about some EXPLAIN ANALYZEs and not just
EXPLAINs?

			regards, tom lane

In response to

• Re: Query optimizer 8.0.1 (and 8.0) at 2005-02-04 23:50:30 from pgsql

Responses

• Re: Query optimizer 8.0.1 (and 8.0) at 2005-02-05 12:29:33 from pgsql
• Re: Query optimizer 8.0.1 (and 8.0) at 2005-02-06 17:29:52 from Ron Mayer

pgsql-hackers by date

Next:	From: Bruce Momjian	Date: 2005-02-05 01:05:15
	Subject: Re: libpq API incompatibility between 7.4 and 8.0
Previous:	From: pgsql	Date: 2005-02-04 23:50:30
	Subject: Re: Query optimizer 8.0.1 (and 8.0)