Is a table with 10 million of rows = with a primary key but conditions used in select are not part of = PK.

My table is = like this:

Column 1 – PK

Column 2 = – Indexed

Column 3 – = Indexed

Column 2 = and Column 3 are Indexed in the same index but they are not mark as = unique or PK

The Query = is like this:

Select column1 from myTable where = Column2 between X and Y

I am = expecting just one record

De: Michael = Holt [mailto:michael@aers.ca]
Enviado el: jueves, 16 de = diciembre de 2010 04:23 p.m.
Para: Anibal David Acosta; = pgsql-admin@postgresql.org
Asunto: RE: [ADMIN] using = limit

This really depends on the type of = query you’re talking about. If there’s only one row in the = table you’re querying then no, I don’t think it’ll = change anything. If you’re querying a single row using a primary = key it shouldn’t change anything. If you’re doing an = aggregate query, say a sum of a bunch of rows, it also won’t = improve performance.

If you’re doing a query on a = table with multiple rows and not filtering by a primary key or other = unique index then yes, it will improve the query.

From:= pgsql-admin-owner@postgr= esql.org [mailto:pgsql-admin-owner@postgr= esql.org] On Behalf Of Anibal David Acosta
Sent: = Thursday, December 16, 2010 11:19 AM
To: pgsql-admin@postgresql.org=
Subject: [ADMIN] using limit

I really like to do efficient SQL = queries so, my question is if I am expecting no more than one row from a = select, using the LIMIT 1 could improve the = performance?

If I use my logic, the LIMIT 1 instruction tell to postgres = that stop searching when found 1 record, but maybe it is = unnecessary

Thanks

Anibal

Yes, a limit should speed up your query from the sounds of things.<= /p>
=A0
From: Anibal David Acosta [mail= to:aa@devshock.com]
Sent: Thursday, December 16, 2010 11:34 AM
To: 'Michae= l Holt'; pgsql-admin@post= gresql.org
Subject: RE: [ADMIN] using limit

=A0
Is a table with 10 million of rows with a primary key but con= ditions used in select are not part of PK.
=A0

My table is like this:=
Column 1 = =96 PK
Colum= n 2 =96 Indexed

Column 3 =96 Indexed
=A0
Column 2 and Column 3= are Indexed in the same index but they are not mark as unique or PK=

=A0
The Query is like this:=
Select column1 fro= m myTable where Column2 between X and Y

=A0
I am expecting just one record=
=A0<= /p>
=A0
=A0

De: Michael Hol= t [mailto:michael@aers.ca]
Enviado el: jueves, 16 de diciembre de 2010 04:23 p.m.
Para: Anibal David Acosta; pgsql= -admin@postgresql.org
Asunto: RE: [ADMIN] using limit<= /p>

=A0

This really depends on the = type of query you=92re talking about. If there=92s only one row in the tabl= e you=92re querying then no, I don=92t think it=92ll change anything. If yo= u=92re querying a single row using a primary key it shouldn=92t change anyt= hing. If you=92re doing an aggregate query, say a sum of a bunch of rows, i= t also won=92t improve performance.

=A0
If you=92re doing a query on a= table with multiple rows and not filtering by a primary key or other uniqu= e index then yes, it will improve the query.

=A0
From: pgsql-admin-owner@postgre= sql.org [mailto:pgs= ql-admin-owner@postgresql.org] On Behalf Of Anibal David Acosta<= br> Sent: Thursday, December 16, 2010 11:19 AM
To: pgsql-admin@postgresql.org
Sub= ject: [ADMIN] using limit

I really like to= do efficient SQL queries so, my question is if I am expecting no more than= one row from a select, using=A0 the LIMIT 1 could improve the performance?=