Re: Compare rows - Mailing list pgsql-performance

From greg@turnstep.com
Subject Re: Compare rows
Date
Msg-id 1756cdfe0672f1cab443627e1fd7a4b3@biglumber.com
Whole thread Raw
In response to Re: Compare rows  (Josh Berkus <josh@agliodbs.com>)
List pgsql-performance
-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1


> The most efficient way for you to store data would be like this:
> main table
> id  address
> 3   67.92
> 7   69.5
>
> child table
> id  value_type  value
> 3   uptime      0.3
> 3   memory      37
> 7   uptime      1.1
> 7   memory      15

Actually, a more efficient* way is this:

value table
vid value_name
1  uptime
2  memory

child table
id   vid   value
3     1     0.3
3     2     37
7     1     1.1
7     2     15


* Still not necessarily the *most* efficient, depending on how the
values are distributed, but it sure beats storing "uptime" over
and over again. :)


- --
Greg Sabino Mullane greg@turnstep.com
PGP Key: 0x14964AC8 200310101243
-----BEGIN PGP SIGNATURE-----
Comment: http://www.turnstep.com/pgp.html

iD8DBQE/huHxvJuQZxSWSsgRAiMNAKD4kQCwdv3fXyEFUu64mymtf567dwCcCKd5
ZzJaV7wjfs00DBT62bVpHhs=
=32b8
-----END PGP SIGNATURE-----



pgsql-performance by date:

Previous
From: Bill Moran
Date:
Subject: Re: One or more processor ?
Next
From: Jeff
Date:
Subject: Re: PostgreSQL Scalable ?