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: DB2 HADB

Re: DB2 HADB

From: DA Morgan <damorgan_at_psoug.org>
Date: Fri, 25 Nov 2005 15:05:53 -0800
Message-ID: <1132959947.306333@yasure>


Mark A wrote:

> "Mark A" <nobody_at_nowhere.com> wrote in message 
> news:z5udnftJ7ZzZDxrenZ2dnUVZ_sCdnZ2d_at_comcast.com...
> 

>>The main reason is that DB2 has table clustering, which Oracle does not
>>have unless you use an index ordered table (which maintains the data in
>>exact order all of the time like an index, and is can be a bit
>>time-consuming for inserts due to the extra work involved).
>>
>>DB2 clustering only "tries" to insert the row on the correct page (never
>>rearranges rows on the page like an index), and if the correct page or a
>>nearby page is not available, then it doesn't waste time with page splits
>>(just puts the row anywhere). The area on a page used for inserting new
>>rows is different on Oracle compared to DB2, and a reorg restores the free
>>space percent (as specified when the table is created) on each page for
>>new inserts in when a clustering index is defined.
>>
>>DB2 table clustering is not "needed", and it is optional, but in many
>>cases it does provide a definite performance enhancement with the extra
>>overhead of an index ordered table in Oracle. Since DB2 has true on-line
>>reorgs, there is no downtime, so I don't know it bothers you so much.
>>
> 
> 
> Correction:
> 
> it does provide a definite performance enhancement WITHOUT the extra 
> overhead
> of an index ordered table in Oracle. 

Wrong again! Thanks for the encore.

-- 
Daniel A. Morgan
http://www.psoug.org
damorgan_at_x.washington.edu
(replace x with u to respond)
Received on Fri Nov 25 2005 - 17:05:53 CST

Original text of this message

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