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: Oracle & RAID strategy question

Re: Oracle & RAID strategy question

From: Hieraklion <hieraklion_at_spray.fr>
Date: Thu, 19 Apr 2001 15:23:03 +0200
Message-ID: <3ADEE6B7.E85EC4E6@spray.fr>

"Bonjour" from Paris,

Not really. It seems to be relative to Quick I/O or not. From a Veritas "white paper" on Veritas Dabase Edition (i.e using volume manager and Vxfs fs) :

The db block size must be a multiple of the fs block size. For db supporting online transaction processing workloads, TPC-C benchmark tests provide evidence that the optimal Oracle db block size is 2K for Quick I/O and 8K for buffered or direct I/0. Because the fs block size does not seem to affect the db performance, using the default size is recommended. The default fs block size varies with the overall size of the fs.

Nuno Souto a écrit :

> On Tue, 17 Apr 2001 14:30:30 +0200, Hieraklion <hieraklion_at_spray.fr>
> wrote:
>
> >
> >For moast workloads, use the default 64K stripe unit size for stripped
> >volumes and 16K for Raid-5 volumes.
> >
>
> Thanks for the info. Intersting. I wonder if there is any
> recomendation is what regards db block size for this stripe size?
>
> Cheers
> Nuno Souto
> nsouto_at_bigpond.net.au.nospam
> http://www.users.bigpond.net.au/the_Den/index.html
Received on Thu Apr 19 2001 - 08:23:03 CDT

Original text of this message

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