[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

RE: thanks (fwd)

This archive contains answers to questions sent to Unidata support through mid-2025. Note that the archive is no longer being updated. We provide the archive for reference; many of the answers presented here remain technically correct, even if somewhat outdated. For the most up-to-date information on the use of NSF Unidata software and data services, please consult the Software Documentation first.

  • Subject: RE: thanks (fwd)
  • Date: Fri, 29 Oct 1999 09:11:49 -0600 (MDT)

===============================================================================
Robb Kambic                                Unidata Program Center
Software Engineer III                      Univ. Corp for Atmospheric Research
address@hidden             WWW: http://www.unidata.ucar.edu/
===============================================================================

---------- Forwarded message ----------
Date: Fri, 29 Oct 1999 09:08:22 -0600
From: Ben Domenico <address@hidden>
To: Basil L. Irwin <address@hidden>
Subject: RE: thanks

Basil,

The Web100 workshop was very valuable to me.  I think it's very important
for us to stay in touch with the high performance networking community
represented there.  At the meeting I made arrangements with two of the
network performance measurements tools groups (Todd Hansen with AMP,
http://amp.nlanr.net and Andy Adams of Carnegie Mellon
http://www.psc.edu/~akadams) to follow up on how we might work together to
bring their tools into our community.  Also, Russ and I agreed that it makes
sense for one of us to give a presentation about the Unidata IDD
(http://www.unidata.ucar.edu/projects/idd/index.html) at the December NLANR
workshop in Miami.  In terms of your major focus -- the tuning issues -- we
may just have to wait to see what you develop.  Because of the wide
differences among our community members, we tend to stay away from mods to
things like TCP and operating systems.  For example Russ reminded me that
our LDM works at the remote procedure call level rather than sockets so it's
tricky to try even simple increases in TCP window sizes for sites on fat
pipes.  However, that's not to say that we aren't interested in your
performance tuning efforts  and  of course we are perfectly willing to take
advantage of it after you've done all the work.  (smile face here)
Seriously though, there may be a place for Unidata in the early user
community and in providing input as to user needs -- especially where you
are discussing APIs for tuning options.

I was going to write up a summary for our techie staff, but it looks like
this message actually is a quick summary of the highlights from my
perspective, so I'm going to send a copy to them.  Please send me a pointer
to your meeting summary when it's done, and I'll forward it to the staff
here.

I hope my naivete didn't slow things down or divert the discussion too much.

Thanks for including us.

-- Ben

-----Original Message-----
From: Basil L. Irwin [mailto:address@hidden]
Sent: Friday, October 29, 1999 8:18 AM
To: address@hidden
Subject: thanks


Ben, I hope the Workshop was of some value to you. Also, I hope you'll
be thinking some more about the value of our work to Unidata, and how we
might work together on these issues.  basil.