Re: enable_nestloop - Mailing list pgsql-general

From Tom Lane
Subject Re: enable_nestloop
Date
Msg-id 9881.1006546392@sss.pgh.pa.us
Whole thread Raw
In response to enable_nestloop  (Luis Amigo <lamigo@atc.unican.es>)
List pgsql-general
Luis Amigo <lamigo@atc.unican.es> writes:
> if we use set enable_nestloop=off the result is:

> Limit  (cost=100160926.76..100160926.76 rows=100 width=120)
>   ->  Sort  (cost=100160926.76..100160926.76 rows=100 width=120)
>         ->  Nested Loop  (cost=100160637.42..100160923.43 rows=100
> width=120)

It's choosing a nestloop at the top level despite the large penalty
added to the cost, presumably because it cannot find any other way to
do the top-level join.  You won't get any more useful answer than that
as long as you're not showing us the query and table schemas involved...

            regards, tom lane

pgsql-general by date:

Previous
From: Doug McNaught
Date:
Subject: Re: explain analyze
Next
From: Antonio Sergio de Mello e Souza
Date:
Subject: Re: newbie needs help with postgres internal tables