Re: ORION results
From: Jared Still <jkstill_at_gmail.com>
Date: Fri, 29 Jan 2010 10:16:57 -0800
Message-ID: <bf46381001291016y6ac459b2o689af26488fe29a_at_mail.gmail.com>
On Fri, Jan 29, 2010 at 8:10 AM, D'Hooge Freek <Freek.DHooge_at_uptime.be>wrote:
Date: Fri, 29 Jan 2010 10:16:57 -0800
Message-ID: <bf46381001291016y6ac459b2o689af26488fe29a_at_mail.gmail.com>
On Fri, Jan 29, 2010 at 8:10 AM, D'Hooge Freek <Freek.DHooge_at_uptime.be>wrote:
> I think block devices are buffered, so you would need to use raw devices
> (or open de block devices with the O_DIRECT flag).
>
>
... or use a benchmark that will obliterate the cache and let you see with
real IO of the storage.
I did that once a few years ago and graphed it - it was quite interesting to
see IO's come down
and disk utilization go up and meet at a point predicted by reading the spec
sheet of the drives.
Unfortunately I seem to have lost that file with the graph in it.
Jared Still
Certifiable Oracle DBA and Part Time Perl Evangelist
Oracle Blog: http://jkstill.blogspot.com
Home Page: http://jaredstill.com
-- http://www.freelists.org/webpage/oracle-lReceived on Fri Jan 29 2010 - 12:16:57 CST