Re: ORION num_disks
From: Greg Rahn <greg_at_structureddata.org>
Date: Fri, 12 Sep 2008 13:17:11 -0700
Message-ID: <a9c093440809121317o149d3acdh221fa75a54d588a3@mail.gmail.com>
Date: Fri, 12 Sep 2008 13:17:11 -0700
Message-ID: <a9c093440809121317o149d3acdh221fa75a54d588a3@mail.gmail.com>
I believe num_disks has to do with the number of I/O threads that are
spawned and num_large has to do with the number of outstanding I/Os
that are targeted to be issued.
For what I use the tool for (I/O bandwidth testing) I generally run the sequential workload to get a best possible data point and then use the rand workload to get numbers closer to what a PQ workload would be.
On Fri, Sep 12, 2008 at 10:47 AM, Allen, Brandon
<Brandon.Allen_at_oneneck.com> wrote:
> In the case of an advanced run like you have below with a specific point
> of 45 large IOs and 0 small IOs, I don't think the num_disks parameter
> does anything, but please let me know if I'm wrong.
-- Regards, Greg Rahn http://structureddata.org -- http://www.freelists.org/webpage/oracle-lReceived on Fri Sep 12 2008 - 15:17:11 CDT