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

Home -> Community -> Mailing Lists -> Oracle-L -> Re: Why "Separating Data and Indexes improves performance" is a myth?

Re: Why "Separating Data and Indexes improves performance" is a myth?

From: Daniel Fink <Daniel.Fink_at_Sun.COM>
Date: Fri, 23 Apr 2004 13:41:52 -0600
Message-id: <40897180.9F5F30B9@sun.com>


Rollback/Undo segments are not written from beginning to end. Blocks in an undo segment are reused, perhaps by the original segment, perhaps 'stolen' by another segment or deallocated. I've never done any formal testing (if anyone has, please correct me), but it is logical to me that undo i/o patterns are similar to data/index segments. It would not suprise me if the i/o pattern of temp tablespaces are closer to data/index/undo than redo.

Redo is written from beginning to end.

Daniel Fink

"Nelson, Allan" wrote:

> You should also separate rollback segments and temp space as they have
> markedly different use patterns as opposed to data and index segments.
> Rollback segments tend to be written from beginning to end and over
> again and temp also tends to be sequential type i/o
>
> Allan



Please see the official ORACLE-L FAQ: http://www.orafaq.com

To unsubscribe send email to: oracle-l-request_at_freelists.org put 'unsubscribe' in the subject line.
--
Archives are at http://www.freelists.org/archives/oracle-l/
FAQ is at http://www.freelists.org/help/fom-serve/cache/1.html
-----------------------------------------------------------------
Received on Fri Apr 23 2004 - 15:12:56 CDT

Original text of this message

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