CURRENT_MEETING_REPORT_


Reported by Jim McQuaid/Wandel & Goltermann Technologies

Minutes of the Benchmarking Methodology Working Group (BMWG)

The meeting included discussion of revision of the current documents and
future work.

The current documents being considered are the ``Benchmarking
Methodologies for Network Interconnect Devices'' (methods) and the
``Benchmarking Methodologies:  Test Frame Formats'' (frames) currently
in the archive.

Several general structural issues were considered first.  The methods
document refers to an Appendix A and B which are currently missing.  It
was agreed to keep these in.  (Appendix A is advice about specific
testing approaches, and B is a table of maximum frame rates for selected
frame sizes on four LANs.)  Scott Bradner agreed to supply a draft of
each.  In addition, a number of revisions and suggestions will result in
an enlargement of Appendix A.

It was also agreed to incorporate the frames document into the methods
document as Appendix C, with some sections being inserted directly into
the relevant section of the methods document.  The frame examples will
be limited to TCP/IP over Ethernet only as a template.

There were many small details of the document and the method which were
discussed and altered during the meeting.  Two ``new'' issues surfaced
as a result of discussions and clarification of some sections:


  1. LAN to WAN test scenarios; and
  2. Routing update performance testing.


Kevin Dubray agreed to supply initial language addressing LAN to WAN
test scenarios, and it became clear that additional expertise on routing
protocol performance and behavior was required to make progress on
routing update and performance testing.

The edited draft will be placed in the working group archives and a
notice will be sent to the list requesting detailed reviews.

Two future topics facing the working group, as stated in the charter,
were briefly discussed at the end of the meeting.  These are host
implementation/protocol performance testing methods and overall network
performance testing methods.  Those present expressed more interest in
the latter, which may be taken up first.


Attendees

Bashir Ashrafi           bashraf@chipcom.com
Stephen Batsell          batsell@itd.nrl.navy.mil
Scott Bradner            sob@harvard.edu
Greg Celmainis           gregc@newbridge.com
Kevin Dubray             kdubray@wellfleet.com
Jeff Hughes              jeff@col.hp.com
Robin Iddon              robini@cix.compulink.co.uk
Merike Kaeo              mkaeo@cisco.com
Edwin King               eek@atc.boeing.com
Jim McQuaid              mcquaid@wg.com
Gerry Meyer              gerry@spider.co.uk
Donald Pace              pace@cntfl.com.
K. K. Ramakrishnan       rama@erlang.enet.dec.com
Paul Serice              serice@cos.com
Mark Swanson             marks@ngc.com
Jerry Toporek            jt@mentat.com
Chris Wellens            chrisw@netcom.com
Cathy Wittbrodt          cjw@barrnet.net