Re: Storage array advice anyone?
- From: "Jonathan Lewis" <jonathan@xxxxxxxxxxxxxxxxxx>
- To: <oracle-l@xxxxxxxxxxxxx>
- Date: Fri, 17 Dec 2004 15:34:04 -0000
No. But I am implying that you think very carefully
about the amount of data that you read/write to a
single disk before you move on to the next disk.
(There are many different terms for this, which is
why I chose to describe the mechanics rather than
give it a name).
Some people set their systems to 32K per disc
at the SAN, then get Oracle to do 1MB direct
reads and wonder why things don't work well.
It seems that many suppliers can't go higher than
1MB per disc (at least, whenever I've been on
site asking what the maximum is, I've invariably
been told that it's 1MB). This means a typical
1MB read will activate two discs - but since many
S/As then decide to put a filesystem stripe on top
of the SAN stripe, who knows what else can go
wrong.
Regards
Jonathan Lewis
http://www.jlcomp.demon.co.uk/faq/ind_faq.html
The Co-operative Oracle Users' FAQ
http://www.jlcomp.demon.co.uk/seminar.html
Optimising Oracle Seminar - schedule updated Sept 19th
----- Original Message -----
From: "Hameed, Amir" <Amir.Hameed@xxxxxxxxx>
To: <jonathan@xxxxxxxxxxxxxxxxxx>
Sent: Friday, December 17, 2004 1:37 PM
Subject: RE: Storage array advice anyone?
Hi Jonathan,
Are you suggesting that there is no benefit in having a larger stripe
width in comparison to a smaller one, for example, 8-way versus 4-way ?
Thanks
-----Original Message-----
From: oracle-l-bounce@xxxxxxxxxxxxx
[mailto:oracle-l-bounce@xxxxxxxxxxxxx] On Behalf Of Jonathan Lewis
Sent: Friday, December 17, 2004 4:23 AM
To: Oracle-L
Subject: Re: Storage array advice anyone?
Sorry to come in so late on this one - I've
had a busy three months, and only just got
back to reading the mail.
Personally I find the whole 'virtualization' thing a
complete con-trick.
Sure, I now have a LUN which is really 50
different spindles - so what good is that if I
send a 'single read request' and that activates
eight of them. It only takes 8 requests like
that and there are 64 reads queued up somewhere,
and who knows where they might be ? Ask
Cary Millsap about queueing and unstable
response times. (Then ask Stephen Barr what
the minimum and maximum response times were
for his Parallel Query problem).
--
//www.freelists.org/webpage/oracle-l
Other related posts:
