Oracle FAQ | Your Portal to the Oracle Knowledge Grid |
Home -> Community -> Usenet -> c.d.o.server -> Re: Physical Layout of disk to use Oracle
Brian Peasland wrote:
>>Also Adams himself says: "Indiscriminate broad striping randomizes
>>all I/O and makes service times uniform." And I say - makes service times
>>more uniform and more predictable over the time.
>>
>>
>
>But isn't that playing down the lowest common denominator? We can either
>strive to make every individual the best that they can be, or we can
>make us all the same. Since one of us can't be the cream of the crop,
>none of us are.
>
>Just my 3.14159265 cents worth,
>Brian
>
>
>
Well yes and no. Should we all in an idealized world be tuning every
last SQL statement and monitoring I/O on every physical drive. You
betcha. But in the real world ... if SAME is good enough then where is
the benefit to the organization to pay some
DBA to get into all that unnecessary effort.
A SQL statement that returns a result to the user in 0.1 second may be better than one that returns in 0.2 second but so what. No end user is ever going to recognize the difference.
I go toward SAME and raw devices unless and until testing proves it insufficient. Then ... it is still cheaper to buy more disk than to tune. How many hours of tuning equate to one more high speed disk? With the current price of disk ... not very many.
-- Daniel Morgan http://www.outreach.washington.edu/ext/certificates/oad/oad_crs.asp http://www.outreach.washington.edu/ext/certificates/aoa/aoa_crs.asp damorgan_at_x.washington.edu (replace 'x' with a 'u' to reply)Received on Tue Sep 16 2003 - 17:26:32 CDT