HansF wrote:
> On Tue, 06 Dec 2005 15:10:16 -0800, DA Morgan wrote:
>
>
>>Bob Jones wrote:
>>
>>
>>>I have no doubt you have those experiences, but I find it hard to accept
>>>that kind of generalization about SAs, or RAID 5 for that matter.
>>
>>http://www.oracle.com/technology/deploy/availability/htdocs/hafaq.html
>>
>>"RAID 5 is good for read intensive, but not write intensive applications."
>
>
> Which can be interpreted as "RAID 5 is great for OLTP. And potentially
> a total waste of time and $ for Warehouses - which are rebuilt anyway so
> there may be no real need to protect the data."
>
> Of course - it depends ...
>
> Funny thing is, people defend RAID 5 as cheaper than the others, and then
> defend the need for the huge cache in front of RAID 5 for speed and then
> defend RAID 5 as being as good as the others but not quite as inexpensive
> because of the huge cache ... ???
>
> <sigh>
What people that defend RAID 5 can not do, however, is repeal the laws
of physics. Parity calculations and 200%+ more writes can not be free.
--
Daniel A. Morgan
http://www.psoug.org
damorgan_at_x.washington.edu
(replace x with u to respond)
Received on Wed Dec 07 2005 - 11:57:41 CST
