Re: Very slow queries - Mailing list pgsql-performance

From Chad Wagner
Subject Re: Very slow queries
Date
Msg-id 81961ff50701301437t60c5370bvdf165215cbeaeb25@mail.gmail.com
Whole thread Raw
In response to Very slow queries  (Sidar López Cruz <sidarlopez@hotmail.com>)
Responses Re: Very slow queries  (Sidar López Cruz <sidarlopez@hotmail.com>)
List pgsql-performance
On 1/30/07, Sidar López Cruz <sidarlopez@hotmail.com> wrote:
query: Delete From ceroriesgo.salarios Where numero_patrono Not In (Select
numero_patrono From ceroriesgo.patronos)

Seq Scan on salarios  (cost=51021.78..298803854359.95 rows=14240077 width=6)
  Filter: (NOT (subplan))
  SubPlan
    ->  Materialize  (cost=51021.78..69422.58 rows=1032980 width=25)
          ->  Seq Scan on patronos  (cost=0.00..41917.80 rows=1032980
width=25)

How many rows exist in salarios, but not in patronos?  How many rows are there in salarios?

What does the explain look like for:

delete
 from ceroriesgo.salarios s
where not exists (select 1
                            from ceroriesgo.patronos
                          where numero_patrono = s.numero_patrono);

Also, is this not a case for a foreign key with a cascade delete?

http://www.postgresql.org/docs/8.2/static/ddl-constraints.html


--
Chad
http://www.postgresqlforums.com/

pgsql-performance by date:

Previous
From: Ted Allen
Date:
Subject: Re: Very slow queries
Next
From: Josh Berkus
Date:
Subject: Re: Querying distinct values from a large table