From mboxrd@z Thu Jan  1 00:00:00 1970
Content-Type: multipart/mixed; boundary="===============5998816234542985676=="
MIME-Version: 1.0
From: Walker, Benjamin <benjamin.walker at intel.com>
Subject: Re: [SPDK] Performance
Date: Tue, 06 Dec 2016 16:58:04 +0000
Message-ID: <1481043483.4630.51.camel@intel.com>
In-Reply-To: 3E90D03A-FA74-4C9F-96E0-0E77A32BA29C@intel.com
List-ID: <spdk@lists.01.org>
To: spdk@lists.01.org

--===============5998816234542985676==
Content-Type: text/plain; charset="utf-8"
MIME-Version: 1.0
Content-Transfer-Encoding: quoted-printable

On Tue, 2016-12-06 at 16:35 +0000, Harris, James R wrote:
> =

> > On Dec 6, 2016, at 5:49 AM, Kumaraparameshwaran Rathnavel <krath(a)clou=
dsimple
> > .com> wrote:
> > =

> > The setup is the default setup of NVMf target with Direct attached
> > Subsystem. The FIO plugin was at the target side and the IOPS from SPDK=
 was
> > found to be around 2.5K and the IOPS from kernel was found to be 100k.
> =

> =

> Hi Param,
> =

> Can you check the AcceptorPollRate setting in your configuration file? =
=C2=A0There
> was a recent fix to the default value for this parameter. =C2=A0You shoul=
d set it
> to at least 10000 (10ms).
> =

> https://github.com/spdk/spdk/commit/5f3761cf7ec50cf03183203dc50657a9f2a18=
908
> =

> Thanks,

Just to reinforce here - this is the solution. We saw this internally a few
weeks ago. Increasing the AcceptorPollRate to at least 10ms (you can go as =
much
as 1 second if you want, but the performance benefit decreases the larger y=
ou
make it), or moving the accceptor to a different core using the AcceptorCore
option will fix the problem.

> =

> -Jim
> =

> > =

> > Regards,
> > Param.
> > > On 06-Dec-2016, at 5:57 PM, Andrey Kuzmin <andrey.v.kuzmin(a)gmail.co=
m>
> > > wrote:
> > > =

> > > =

> > > =

> > > On Tue, Dec 6, 2016, 15:24 Kumaraparameshwaran Rathnavel <krath(a)clo=
udsimpl
> > > e.com> wrote:
> > > > Hi All,
> > > > =

> > > > Is there any benchmarking done to prove that SPDK is better than the
> > > > Kernel driver?
> > > > =

> > > > We did run the sample driver provided from the SPDK and compared wi=
th
> > > > the kernel interface and found that SPDK is very slow. We did run F=
IO.
> > > > =

> > > =

> > > It's worth publishing your setup and results first.
> > > =

> > > Regards,
> > > Andrey
> > > > =

> > > > Can you describe the setup for the performance measurement between
> > > > Kernel and SPDK driver.
> > > > =

> > > > Regards,
> > > > Param.
> > > > _______________________________________________
> > > > SPDK mailing list
> > > > SPDK(a)lists.01.org
> > > > https://lists.01.org/mailman/listinfo/spdk
> > > > =

> > > --=C2=A0
> > > Regards,
> > > Andrey
> > > _______________________________________________
> > > SPDK mailing list
> > > SPDK(a)lists.01.org
> > > https://lists.01.org/mailman/listinfo/spdk
> > =

> > _______________________________________________
> > SPDK mailing list
> > SPDK(a)lists.01.org
> > https://lists.01.org/mailman/listinfo/spdk
> =

> _______________________________________________
> SPDK mailing list
> SPDK(a)lists.01.org
> https://lists.01.org/mailman/listinfo/spdk

--===============5998816234542985676==--