> Hmmm ... why is the inner Nested Loop estimated to produce > 30120 rows, > when the cartesian product of its inputs would only be 285 > * 14 = 3990 > rows? Oh my... didn't notice it!!! > What PG version is this That was 8.4 beta1; now tried on select version() "PostgreSQL 8.4.0 on sparc-sun-solaris2.10, compiled by cc: Sun C 5.9 SunOS_sparc Patch 124867-02 2007/11/27, 64-bit" And got same thing: http://explain-analyze.info/query_plans/3820-query-plan-2526 , and can you extract a > self-contained > test case with an equally silly estimate? mmh... I'll try to "squeeze" the data I'm using and see if it still gets that plan. How big a test case could be? Could I post it (ftp?) somewhere? (you can email me about this in private if you want) -- Sent via pgsql-general mailing list (pgsql-general@xxxxxxxxxxxxxx) To make changes to your subscription: http://www.postgresql.org/mailpref/pgsql-general