Re: [HACKERS] Bug on complex join - Mailing list pgsql-hackers

From Vadim Mikheev
Subject Re: [HACKERS] Bug on complex join
Date
Msg-id 36E5D6DD.749189C7@krs.ru
Whole thread Raw
In response to Re: [HACKERS] Bug on complex join  (Oleg Broytmann <phd@sun.med.ru>)
List pgsql-hackers
Oleg Broytmann wrote:
> 
> Hi!
> 
> On Tue, 9 Mar 1999, Hannu Krosing wrote:
> > > SELECT cn.date_i, cn.pos_id
> > >    FROM central cn
> > >       WHERE cn.date_i >= current_date - '300 days'::timespan
> > >
> > > returns 3156 rows. But this:
> > >
> > > SELECT cn.date_i, p.subsec_id, cn.pos_id, p.pos_id
> > >    FROM central cn, shops sh, districts d, positions p
> > >       WHERE cn.date_i >= current_date - '300 days'::timespan
> >
> > this should return
> >    3156 * count(shops) * count(districts) * count(positions)
> >
> > which is probably too much for the backend ;(                              ^^^^^^^^^^^^^^^
For the client-side, not for the backend - backend doesn't
keep all result tuples in memory.

> 
>    Bad news. Thanks for pointing this.

Vadim


pgsql-hackers by date:

Previous
From: Bruce Momjian
Date:
Subject: Re: [HACKERS] Re: map
Next
From: "D'Arcy" "J.M." Cain
Date:
Subject: Re: [HACKERS] Re: map