Date: Tue, 7 Jan 2003 14:21:19 -0200 (GMT+2)
From: Achilleus Mantzios <achill@matrix.gatewaynet.com>
To: Tomasz Myrta <jasiek@klaster.net>
Cc: pgsql-performance@postgresql.org, <pgsql-sql@postgresql.org>,
	<pgsql-general@postgresql.org>
Subject: Re: [SQL] 7.3.1 index use / performance
In-Reply-To: <3E1AC14C.7010104@klaster.net>
Message-ID: <Pine.LNX.4.44.0301071410270.8183-100000@matrix.gatewaynet.com>
MIME-Version: 1.0
Content-Type: TEXT/PLAIN; charset=US-ASCII

On Tue, 7 Jan 2003, Tomasz Myrta wrote:

> Maybe it is not an answer to your question, but why don't you help
> Postgres by yourself?

Thanx,

i dont think that the issue here is to help postgresql by myself.
I can always stick to 7.2.3, or use indexes that 7.3.1 will
acknowledge, like noonf_vcode_date on noon (v_code,report_date).
(unfortunately when i create the above noonf_vcode_date index, it is only
used until
the next vacuum analyze, hackers is this an issue too???),
but these options are not interesting from a postgresql perspective :)

> For this kind of queries it's better to drop index on report_date - your
> report period is one year and answer to this condition is 10% records (I
> suppose)

I cannot drop the index on the report_date since a lot of other queries
need it.

> It would be better to change 2 indexes on v_code and rotation into one
> index based on both fields.
> What kind of queries do you have? How many records returns each "where"
> condition? Use indexes on fields, on which condition result in smallest
> amount of rows.
>
> Regards,
> Tomasz Myrta
>

==================================================================
Achilleus Mantzios
S/W Engineer
IT dept
Dynacom Tankers Mngmt
Nikis 4, Glyfada
Athens 16610
Greece
tel:    +30-10-8981112
fax:    +30-10-8981877
email:  achill@matrix.gatewaynet.com
        mantzios@softlab.ece.ntua.gr