Re: single session I/O bandwidth

From: joshuasingham <joshuasingham_at_gmail.com>
Date: Tue, 12 Mar 2013 19:38:41 +0800
Message-ID: <CAKYyGsG19+7cv_rWa_p7uXh6SULFaQ80_-1V3vO=AY-YqWFvdA_at_mail.gmail.com>

I believe the single cpu thread is limiting the data consumption to cpu(you will be able to get the answer from the strace or dtrace mention in this email thread) and also have you set the DB_FILE_MULTIBLOCK_READ_COUNT to a small value that could also limit the throughout . Regards

Joshua

follow me on

https://twitter.com/joshuasingham

On Tue, Mar 12, 2013 at 5:59 PM, Nuno Souto <dbvision_at_iinet.net.au> wrote:

> and if you use dd, don't forget to set bs, ibs or obs to 8K or whatever
> the db block size is..
> Otherwise you'll be measuring times for 512 byte I/O.
>
> --
> Cheers
> Nuno Souto
> dbvision_at_iinet.net.au
>
>
>
> On 12/03/2013 6:05 PM, Iggy Fernandez wrote:
> > Why is the basis of the assumption that a single Oracle thread has more
> capacity that that? Trace one of the threads for a while using "strace -c"
> to produce a summary and you will see how time is spent within the thread.
> > You can also check a summary trace of a "dd of=/dev/null" command.
> Chances are you will see higher throughput but dd much less processing than
> an Oracle thread.
> > Kindest regards,
> >
>
> --
> http://www.freelists.org/webpage/oracle-l
>
>
>

--
http://www.freelists.org/webpage/oracle-l

Received on Tue Mar 12 2013 - 12:38:41 CET