Oracle FAQ Your Portal to the Oracle Knowledge Grid
HOME | ASK QUESTION | ADD INFO | SEARCH | E-MAIL US
 

Home -> Community -> Usenet -> c.d.o.server -> Re: Designing Large tables

Re: Designing Large tables

From: Clive Cooper <clic_at_ACTRIX.GEN.NZ>
Date: 1997/02/07
Message-ID: <32FAE389.53A1@ACTRIX.GEN.NZ>#1/1

Roger Wernersson wrote:
>
> Lun Wing San (Oracle) wrote:
> >
> > Hitesh Patel wrote:
> > >
> > > We are in the process of creating an archive database which consists
> > > of few tables of sizes ranging from 4 Gb to 16 Gbytes. Do you have
> > > any suggestions for creating these tables in terms of partitioning the
> > > tables, spreading across disks, accessing the data, clustering etc.
> > > Does anyone has this kind of scenario or know of any articles on the
> > > internet ?
> >
> > You can try to use partition view to implement it. By using partition
> > views, you will gain several advantages in performance.
>
> This sounds interesting but what is a partition view?
>
> --
> Sport radio: people listening to people watching people having fun
> Mailto:roger.wernersson_at_adra.se
> BTW: All opinions are mine, all mine, and nobody's but mine.
Partition views are described in Oracle7 Server Tuning manual appendix C.

That manual also has a chapter on tuning i/o.

The partitioned view approach helps with overall management of the space. You still need to tune the i/o for performance. With tables of several million rows performance is of great importance for users of the archive.

I manage a multi gigabyte datawarehouse using Oracle and have spent a lot of time (and the company money :) ) on the i/o.

Good luck. Received on Fri Feb 07 1997 - 00:00:00 CST

Original text of this message

HOME | ASK QUESTION | ADD INFO | SEARCH | E-MAIL US