Laurette Cisneros <laurette(at)nextbus(dot)com> writes:
> Will keep trying. I also noticed that several triggers were dropped when I
> dropped the table. I need to add these back and see if they make a
> difference.
No, join_selectivity isn't going to care about triggers. What's failing
is the estimation of the fraction of rows that will match on address
between the two tables (join_selectivity is rejecting the result as
obviously bogus, which it is). That doesn't depend on anything except
the ANALYZE statistics.
regards, tom lane
In response to
pgsql-hackers by date
| Next: | From: Bruce Momjian | Date: 2002-01-02 23:17:42 |
| Subject: Re: software license question |
| Previous: | From: Laurette Cisneros | Date: 2002-01-02 22:32:52 |
| Subject: Re: bug in join? |