Re: [bmwg] Call for Adoption: draft-morton-bmwg-b2b-frame

["MORTON, ALFRED C (AL)" <[e-mail]>]

Hi Yoshiaki Itoh,

Thank you for your past reviews and most recent comments.

The main innovations of this draft:

                  https://datatracker.ietf.org/doc/draft-morton-bmwg-b2b-frame/<https://urldefense.proofpoint.com/v2/url?u=https-3A__datatracker.ietf.org_doc_draft-2Dmorton-2Dbmwg-2Db2b-2Dframe_&d=DwMFAg&c=LFYZ-o9_HUMeMTSQicvjIg&r=OfsSu8kTIltVyD1oL72cBw&m=HS7Ed6lrlBb60NZsCWT7WGKCJJIeGwqkssw7Jo0fsCk&s=Oqg5iKz3YLnA2seSHeJEgwf-Ga7nAK8GargCnf52jKM&e=>
(which is currently considered for WG adoption in BWMG,
and comments on this call are appreciated)

involve the pre-requisite testing for RFC 2544 Throughput
that results in:


1.  development of a correction factor for the calculation

to remove the frames that were forwarded from the frames

that were buffered, and give an accurate estimate of

buffer size for the single port-pair case.

2.  reduction of the back-to-back frame tests needed,

where only packet rates that induce buffering would

be tested.

I shared test results that illustrate the value of these two
innovations:
https://wiki.opnfv.org/display/vsperf/Traffic+Generator+Testing#TrafficGeneratorTesting-AppendixB:Back2BackTestingTimeSeries(fromCI)
and
https://datatracker.ietf.org/meeting/100/materials/slides-100-bmwg-back-to-back-frame-benchmark-01

RFC2544 does not discuss either of these points, so
this draft would update and expand the current test
in the simplest way possible. However, the current draft
benefits from your previous comments, specifically that
we distinguish the scope of the back2back test from the
RFC8239 multi-port testing and cases where line-rate
switching is possible at the smallest frame sizes
(both are out of scope, but there are still many places
where the current back2back frame tests will be useful).

The Pause method you are proposing has some interesting
possibilities for new test methods, in the same scope of the multi-port
scenarios addressed in RFC 8239 (Datacenter).  We discussed this
during our IETF-103 session. I think the WG really appreciates
that you have shared your test results. There were several issues
raised during the session, and further discussion would
benefit from development of a procedure and other details
usually found in an Internet Draft, such as new test equipment
capabilities (ability to send "Pause"). I can help you with
this documentation aspect, if you would like.

It seems clear that we need a set of buffer measurement solutions
that depend on the specific circumstances and scenarios to
measure.

Thanks again for reviewing the back-to-back frame draft,
and for your contributions to the many WG discussions on
buffer measurement. I hope you'll support my draft's first steps
to better measurement, with many steps to follow.

best regards,
Al
(as a participant/author)

From: bmwg [mailto:[e-mail]] On Behalf Of Yoshiaki Itou
Sent: Friday, December 14, 2018 1:05 AM
To: Sarah B <[e-mail]>; [e-mail]
Subject: Re: [bmwg] Call for Adoption: draft-morton-bmwg-b2b-frame

Hello BMWG,

I would like to comment on the definition of back-to-back Frame value.

>5.2.  Test for a Single Frame Size
>
>The Back-to-back Frame value is the longest burst of frames that the
>DUT can successfully process and buffer without frame loss,

The description of this back-to-back Frame value is exact.
To explain more strictly:
The Back-to-back Frame value is the longest burst of frames that the
DUT can successfully process(forward not buffered) and buffer without frame loss,

"Pause" method is a measuring the number of buffered frames only.

https://datatracker.ietf.org/doc/minutes-103-bmwg/<https://urldefense.proofpoint.com/v2/url?u=https-3A__datatracker.ietf.org_doc_minutes-2D103-2Dbmwg_&d=DwQFAg&c=LFYZ-o9_HUMeMTSQicvjIg&r=OfsSu8kTIltVyD1oL72cBw&m=HS7Ed6lrlBb60NZsCWT7WGKCJJIeGwqkssw7Jo0fsCk&s=MjJaOUmICdbkd2ls5SZN47QrTvYrHR76s8UZzrPbm4I&e=>

8. New Buffer assessment method for RFC 8239 Data Center Benchmarking
   Author: Yoshiaki Itou
   Mail List References:
   https://mailarchive.ietf.org/arch/msg/bmwg/lKiImpq8RlNapD8CVRG1dRZlMZ8<https://urldefense.proofpoint.com/v2/url?u=https-3A__mailarchive.ietf.org_arch_msg_bmwg_lKiImpq8RlNapD8CVRG1dRZlMZ8&d=DwQFAg&c=LFYZ-o9_HUMeMTSQicvjIg&r=OfsSu8kTIltVyD1oL72cBw&m=HS7Ed6lrlBb60NZsCWT7WGKCJJIeGwqkssw7Jo0fsCk&s=sloQPaNgjEx7RZCURpl7ReYqeT2eytL6Rr1xQdyq_50&e=>
   https://mailarchive.ietf.org/arch/msg/bmwg/elCgGpsB-TH1zCwaRhzM7B2mW4g<https://urldefense.proofpoint.com/v2/url?u=https-3A__mailarchive.ietf.org_arch_msg_bmwg_elCgGpsB-2DTH1zCwaRhzM7B2mW4g&d=DwMFAg&c=LFYZ-o9_HUMeMTSQicvjIg&r=OfsSu8kTIltVyD1oL72cBw&m=HS7Ed6lrlBb60NZsCWT7WGKCJJIeGwqkssw7Jo0fsCk&s=EsDXYSsfsoIkrcNSC0fp9qYEcx9kJXOq7pTu2Irbs7M&e=>

Best Regards,
Yoshiaki Itou
From: bmwg <[e-mail]<mailto:[e-mail]>> On Behalf Of Sarah B
Sent: Saturday, December 08, 2018 2:57 AM
To: [e-mail]<mailto:[e-mail]>
Subject: [bmwg] Call for Adoption: draft-morton-bmwg-b2b-frame

Hello BMWG,
                  At IETF103 in Bangkok Al Morton, author of the "back to back framework" draft, presented an update.

                  https://datatracker.ietf.org/doc/draft-morton-bmwg-b2b-frame/<https://urldefense.proofpoint.com/v2/url?u=https-3A__datatracker.ietf.org_doc_draft-2Dmorton-2Dbmwg-2Db2b-2Dframe_&d=DwMFAg&c=LFYZ-o9_HUMeMTSQicvjIg&r=OfsSu8kTIltVyD1oL72cBw&m=HS7Ed6lrlBb60NZsCWT7WGKCJJIeGwqkssw7Jo0fsCk&s=Oqg5iKz3YLnA2seSHeJEgwf-Ga7nAK8GargCnf52jKM&e=>

                  There has been sufficient interest, review, and comments on the draft, and I'd like to call for adoption of this draft by the WG. Please reply to this email before December 21, 2018 and indicate:

(1) whether you support addition of the following milestone to
  the BMWG charter:

date TBD: Updates for the Back-to-back Frame Benchmark in RFC 2544


(2) whether you support the adoption of draft-morton-bmwg-b2b-frame-03
      as the basis document for this milestone

(3) whether you commit to reviewing the WG document if adopted



Thanks for your consideration of the draft!

Kind regards,
Sarah
bmwg co-chair