Message-ID: <435CFBB6.2010303@informatik.hu-berlin.de>
Date: Mon, 24 Oct 2005 17:20:22 +0200
From: Silke Trissl <trissl@informatik.hu-berlin.de>
User-Agent: Mozilla/5.0 (X11; U; Linux i686; en-US; rv:1.7.10) Gecko/20050727
MIME-Version: 1.0
To: =?KOI8-R?Q?=E9=CC=D8=D1_=EB=CF=CE=C0=C8=CF=D7?= <ilyakon@mail.ru>
Cc: pgsql-sql@postgresql.org
Subject: Re: 
References: <E1EU3wZ-000Fgr-00.ilyakon-mail-ru@f21.mail.ru>
In-Reply-To: <E1EU3wZ-000Fgr-00.ilyakon-mail-ru@f21.mail.ru>
Content-Type: text/plain; charset=KOI8-R
Content-Transfer-Encoding: quoted-printable

=E9=CC=D8=D1 =EB=CF=CE=C0=C8=CF=D7 wrote:
>>For example, consider the queres:
>>
>>SELECT * FROM table1 WHERE field1=3D1 AND field2=3D1;
>>
>>SELECT * FROM table1 WHERE field2=3D1 AND field1=3D1;
>>
>>
>>These two queries are logically equivalent. But in all cases the planne=
r generates a query plan that performs field1=3D1 condition, and then fie=
ld2=3D1 condition, as there is a index on field1.
>>
>>Is it possible to instruct the PostgreSQL query planner to perform fiel=
d2=3D1 condition first, and then field1=3D1 condition?
>>
You might get the desired result, if you switch off the index scan:

set ENABLE_INDEXSCAN =3D OFF.

But there is no way to tell Postgres what to use first. Usually the
query planer is quite good, so there is no reason to fiddle around. And
why on earth would you like that. In the end you get the same result.

=09
Hope, that helps

	Silke