moving IOR/mdtest into the I/O community
by Glenn Lockwood
Dear IOR and mdtest users and contributors,
The IOR benchmark has recently seen a revival of development with some
substantial new features being contributed by members of the community. At
present, the most feature-rich and actively maintained fork of the original
IOR is at https://github.com/ior-lanl/ior, but the breadth of the community
participating in that fork has grown beyond its original maintainers at Los
Alamos National Laboratory.
In response to this evolution, the current maintainers of IOR-LANL would
like to move the IOR benchmark, including the mdtest benchmark that was
recently merged into this fork, into an institution-neutral location.
Therefore, we are planning to relocate the official IOR/mdtest to
https://github.com/hpc, which is a GitHub organization with representatives
from academia, government, and industry. Following this relocation, we
will begin laying out a roadmap for IOR/mdtest feature development with the
involvement of any and all interested members of the IOR/mdtest community.
Before we relocate IOR/mdtest and begin feature planning, though, we would
like feedback from IOR/mdtest users and contributors to gauge community
interest and address any questions or concerns. Barring any major
objections, we plan to move IOR/mdtest, archive the LANL fork, and begin
roadmap planning in January 2018.
Sincerely,
The IOR maintainer team
5 years, 7 months
Re: [IO-500] Impact of KPTI on IO-500 performance
by Richard Booth
Dear all
Interesting blog from Julian and reading the thread on the BeeGFS google
group. Clearly it is early days and it will good to see more benchmarks.
You may be interested in a white paper Rosemary Francis has written about
mitigating the impact of the Meltdown and Spectre fixes
https://www.ellexus.com/wp-content/uploads/2018/01/180107-Meltdown-and-Sp...
Kind regards
Richard
On 23 January 2018 at 11:00, <io-500-request(a)vi4io.org> wrote:
>
> ----------------------------------------------------------------------
>
> Message: 1
> Date: Tue, 23 Jan 2018 11:01:38 +0100
> From: Julian Kunkel <juliankunkel(a)googlemail.com>
> To: io-500(a)vi4io.org
> Subject: [IO-500] Impact of KPTI on IO-500 performance
> Message-ID:
> <CAMcTfv42WN-4+WevNGbGx2ghXX9JP6Pe_RwWqC2Z=5i
> HCB-gfQ(a)mail.gmail.com>
> Content-Type: text/plain; charset="UTF-8"
>
> Dear all,
> for DKRZ I did some measurements for the performance impact of KPTI on
> tmpfs:
> http://juliankunkel.blogspot.de/2018/01/impact-of-kpti-on-
> io-500-for-dkrz.html
>
> On our system, there is minimal impact -- I also did a latency
> analysis (linked in the post).
> I'm curious if other people experienced something else.
>
> Regards,
> Julian
>
> --
> http://wr.informatik.uni-hamburg.de/people/julian_kunkel
>
>
> ------------------------------
>
> Message: 2
> Date: Tue, 23 Jan 2018 11:30:59 +0100
> From: Sven Breuner <sven.breuner(a)thinkparq.com>
> To: io-500(a)vi4io.org
> Subject: Re: [IO-500] Impact of KPTI on IO-500 performance
> Message-ID: <53e220c7-4580-3d46-176c-2ee1abeba1d0(a)thinkparq.com>
> Content-Type: text/plain; charset="utf-8"; Format="flowed"
>
> Hi,
>
> thanks for sharing this, Julian.
>
> The impact of KPTI is definitely one of the frequently asked questions for
> us
> these days. (Unfortunately it's one of those things where the application
> performance impact is difficult to predict, but that shouldn't stop us
> from trying.)
>
> Not with the IO-500 benchmark suite, but maybe still interesting for some
> people
> on this list could be the BeeGFS performance comparison with and without
> client-/server-side KPTI on the public BeeGFS user mailing list, provided
> by
> Guido from Zuse Institute:
>
> https://groups.google.com/forum/#!msg/fhgfs-user/0cxGJDPMxZg/37QcgttzAQAJ
>
> Best regards
>
> Sven
>
>
> Julian Kunkel wrote on 23.01.2018 11:01:
> > Dear all,
> > for DKRZ I did some measurements for the performance impact of KPTI on
> tmpfs:
> > http://juliankunkel.blogspot.de/2018/01/impact-of-kpti-on-
> io-500-for-dkrz.html
> >
> > On our system, there is minimal impact -- I also did a latency
> > analysis (linked in the post).
> > I'm curious if other people experienced something else.
> >
> > Regards,
> > Julian
> >
>
> --
> <http://www.beegfs.io>
> Sven Breuner
> /CEO/
> *ThinkParQ GmbH*
> w: www.thinkparq.com <http://www.thinkparq.com> e:
> sven.breuner(a)thinkparq.com
> <mailto:sven.breuner@thinkparq.com>
> <https://twitter.com/BeeGFS> <https://www.linkedin.com/in/svenbreuner>
>
> -------------- next part --------------
> An HTML attachment was scrubbed...
> URL: <http://www.vi4io.org/pipermail/io-500/attachments/
> 20180123/8497a5e9/attachment-0001.html>
>
> ------------------------------
>
> Subject: Digest Footer
>
> _______________________________________________
> IO-500 mailing list
> IO-500(a)vi4io.org
> https://www.vi4io.org/cgi-bin/mailman/listinfo/io-500
>
>
> ------------------------------
>
> End of IO-500 Digest, Vol 13, Issue 1
> *************************************
>
--
Richard Booth
Account Manager, Ellexus Ltd
richard.booth(a)ellexus.com
+44 (0)1223 42 16 46
Ellexus is the I/O profiling company.
www.ellexus.com
*Ellexus Ltd is a limited company registered in England & Wales*
*Company registration no. 07166034*
*Registered address: 198 High Street, Tonbridge, Kent TN9 1BE, UK*
*Operating address: St John's Innovation Centre, Cowley Road, Cambridge CB4
0WS, UK*
5 years, 8 months