From: | Sanjay Arora <skpobox(at)gawab(dot)com> |
---|---|
To: | Postgresql Performance <pgsql-performance(at)postgresql(dot)org> |
Subject: | Re: Hardware upgrade for a high-traffic database |
Date: | 2004-08-11 15:47:10 |
Message-ID: | 1092239230.2277.5.camel@Sewak.Asr.Lan |
Views: | Raw Message | Whole Thread | Download mbox | Resend email |
Thread: | |
Lists: | pgsql-performance |
On Wed, 2004-08-11 at 17:51, Merlin Moncure wrote:
> One thing you might consider is materialized views. Your aggregate
> functions are killing you...try to avoid using them (except min/max on
> an index). Just watch out for mutable functions like now().
>
> http://www.varlena.com/varlena/GeneralBits/Tidbits/matviews.html
>
> An application specific approach is to use triggers to keep the data you
> need in as close to query form as possible...you can reap enormous
> savings particularly if your queries involve 3 or more tables or have
> large aggregate scans.
I thought materialized views support in pgsql was experimental as yet.
Are the pg mat-view code upto production servers? Also, do you have to
delete mat-views before you dump the db or does dump automatically not
dump the mat-views data?
Sanjay.
From | Date | Subject | |
---|---|---|---|
Next Message | Tom Lane | 2004-08-11 16:06:38 | Re: Storing binary data. |
Previous Message | Shridhar Daithankar | 2004-08-11 15:12:50 | Re: Storing binary data. |