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: Myth revisited ...

Re: Myth revisited ...

From: Holger Baer <holger.baer_at_science-computing.de>
Date: Mon, 17 Nov 2003 09:38:31 +0100
Message-ID: <bpa1e8$2cp$1@news.BelWue.DE>


Hans Forbrich wrote:
> At the risk of being shot, drawn and quartered:
>
> I know (and agree with) the fundemental discussion that separating
> indexes and tables into separate tablespaces should not be done for
> performance reasons - in pre-Oracle9i environments!
>
> However, with Oracle9i and it's support for multiple block sizes: Is
> there a possible performance benefit to be obtained by placing the
> tables and [some] indexes in separate tablespaces, IF the tablespaces
> have different blocksizes?
>
> (If this has been previously discussed, please just point me to the
> approximate time frame so I can review the archives.)
>
> /Hans

Well, if it means anything to you, Don Burleson promotes it:

http://www.oracle-training.cc/oracle_tips_block_sizes.htm

(I believe I once read a paper on dbazine.com, also by Don, you might want to search there, too).

Anyway, even Oracle makes use of multiple blocksizes in TPC-Benchmarks, exactly to eliminate the need for an overflow segment as Nuno suggested in his post.

I once asked Tom Kyte on asktom about this, however his answer never got published, so I can't point you there. But he didn't see much point in using multiple blocksizes.

Holger Received on Mon Nov 17 2003 - 02:38:31 CST

Original text of this message

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