Re: order of nested loop
Tom Lane <tgl@sss.pgh.pa.us>
From: Tom Lane <tgl@sss.pgh.pa.us>
To: Joseph Shraibman <jks@selectacast.net>
Cc: pgsql-general@postgreSQL.org
Date: 2003-06-17T19:07:19Z
Lists: pgsql-general
Joseph Shraibman <jks@selectacast.net> writes: > => explain select * from usertable where podkey = 20; > Index Scan using usertable_podkey_key on usertable > (cost=0.00..16019.99 rows=5923 width=625) > => select count(*) from usertable where podkey = 20; > count > ------- > 3 > (1 row) Well, there's our problem :-( I would suggest bumping up the statistics target for usertable.podkey (see ALTER TABLE SET STATISTICS). Since it's only an int4 column you could make the target 100 (instead of the default 10) without much cost. That should give substantially finer-grain detail and hopefully bring this estimate down out of the stratosphere. Re-analyze the table and see what it gets you. regards, tom lane