But it also have to compute checksums and wait for all
disks to complete I/O before returning the result. In
my experience, RAID-5 (EMC-Clariion) was not faster then
Symmetrix configured for RAID 1+0. There are great adds
for RAID-5 at http://www.baarf.com written by Art Kagel and
James Morle, and a great add for RAID-5 on http://www.hotsos.com
called "Configuring Oracle for VLDB" by Cary Millsap. Also worth
reading is an article by Hank Tulis called "Aligning Oracle Blocks
with Hardware Stripe Boundaries" which can also be found on Hotsos
and which gives fairly good understanding of the issues with Oracle
and Raid. You should read those articles and show them to your SUN
representative. If he persists, then take the latest edition of Tom
Kyte's expert 1-on-1 and shove it up his...intestines. Tom's book
was selected for size, not for any other reason. You can use
Encyclopedia Britannica or Yellow Pages for NYC instead.
--
Mladen Gogala
A & E TV Network
Ext. 1216
> -----Original Message-----
> From: Nelson, Allan [mailto:anelson_at_miswaco.com]
> Sent: Wednesday, August 25, 2004 4:14 PM
> To: oracle-l_at_freelists.org
> Subject: RE: Raid5 Vs Raid0+1 -- Raw Vs Solaris 9 Concurrent
> Direct IO UFS
>
>
> Raid 5 does get good read performance because for reads more
> disks get in on the action. 3 in his example vs essentially
> 2 for the raid 0+1 stuff.
----------------------------------------------------------------
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 Wed Aug 25 2004 - 15:54:49 CDT
