Cc: 
Subject: index strategies
Content-type: text/plain
Date: Mon, 10 Feb 2003 10:25 +0000
Mime-version: 1.0
To: pgsql-sql@postgresql.org
From: val@webtribe.net
Message-Id: <20030210102507.84E5D18332@cheetah.webtribe.net>
Content-Transfer-Encoding: quoted-printable

Can someone point me in the direction of some good, useful info=20
re: datamine index strategies?

I have one BIG table (apr :
Attribute |           Type           | Modifier
-----------+--------------------------+----------
 dnis      | text                     |
 cli       | text                     |
 iplogin   | text                     |
 brand     | text                     |
 begintime | timestamp with time zone |
 endtime   | timestamp with time zone |
 ipaddr    | text

Any one of the columns may be used in the where clause of a=20
select statement.=20=20
There are also aggregates used in select statements
(e.g., sum(endtime-begintime) group by iplogin AND count(*)=20
group by iplogin).

I have a small clue (ha!) about indexing but want to know the=20
proper way to attack the indexing strategy.

Any info given is v much appreciated.

Val