[net-perf] Re: [HPR-TAC] Progress on Pipes?

[Michael Van Norman]

To update everybody on progress.  ken and I have been working with Russ 
Hobby to host an E2Epi Network Performance Workshop 
(http://e2epi.internet2.edu/network-perf-wk/index.html) at UCLA the week 
of June 27.  The exact dates are still being worked out.  The agenda 
will probably vary from what is on the web site as they are still 
working on the format, but it will be two days in length.  It looks like 
we will be getting space to handle two people from each HPR connector + 
CENIC (so you can start fighting amongst yourselves now :).  Even though 
the dates are not yet set, I wanted to get this out so you can start 
planning your schedules and working on getting the requirements for 
participation 
(http://e2epi.internet2.edu/network-perf-wk/requirements.html) into place.

I should have all the details by the next HPR-TAC meeting.  My 
expectation is that this workshop will kick-off a full piPEs 
infrastructure for CalREN.

On a related front, Mark and I have been working with the NLANR folks on 
the campus network version of their AMP monitors.  I expect that we will 
get copies of the code real soon now.  I hope to have something to 
present on this effort at the next meeting as well.

/Mike

Mark Boolootian wrote:
> Hi Garrett,
> 
> 
>>What is the status of deploying Pipes in CENIC? Do we have all campuses
>>on board with doing this? Is there a timeline for deployment? Who is
>>the CENIC expert in this area? How much progress has been made so far?
> 
> 
> Mike Van Norman and myself are supposed to be leading the effort to
> promote deployment of performance measurement tools within CENIC.  But,
> sigh, work keeps getting in the way of accomplishing anything useful.
> 
> There is a mailing list associated with this effort:
> 
>   http://lists.cenic.org/mailman/listinfo/network-performance
> 
> but no activity.  I believe Mike was going to draft an email that articulated 
> recommendations for campus baseline configurations.  At a minimum, that would
> include OWAMP and BWCTL.  Running an AMP probe, multicast beacon, and NDT 
> server would be extremeley valuable.
> 
> The CENIC expert would have to be Russ.  Russ has attempted to get
> the ball rolling on more than one occasion, but so far there isn't
> anything to report in the way of progress.  UCSC is planning on OWAMP/BWCTL
> deployment (by summer I hope) and I'm fairly certain both UCLA and Berkeley
> plan on deploying said servers as well.  
> 
> I think I can safely say that OWAMP/BWCTL are the mechanisms of choice
> for performance measurement within CENIC, and that's the direction that
> most campuses are headed (albeit slowly).  The real trick is overcoming
> the current inertial state.  Please join the mailing list above and complain
> about the slack leadership :-)
> 
> mb
> _______________________________________________
> HPR-TAC mailing list
> HPR-TAC at lists.cenic.org
> http://lists.cenic.org/mailman/listinfo/hpr-tac