links on the blink (fwd)

From list-admin@merit.edu Sat Nov 4 15:13:10 1995

  >Date: Sat, 4 Nov 1995 12:48:36 -0800 (PST)
  >From: Michael Dillon <michael@memra.com>
  >X-Sender: michael@okjunc.junction.net
  >To: Hans-Werner Braun <hwb@upeksa.sdsc.edu>
  >cc: Mike <mn@tremere.ios.com>, nathan@netrail.net, nanog@merit.edu,
  > D.Mills@cs.ucl.ac.uk
  >Subject: Re: links on the blink (fwd)
  >In-Reply-To: <199511041859.KAA08467@upeksa.sdsc.edu>
  >Message-ID: <Pine.LNX.3.91.951104123440.12997A-100000@okjunc.junction.net>
  
  >
  >> I will not go into a point by point rebuttal here, even though I
  >> generally do not subscribe to your arguments. I am not planning on
  >> "winning" here, I just want to get the issues on the table and evaluate
  >> the solution space. Just let me ask you, as a customer who fairly
  >> frequently experiences 10% packet loss between major Internet locations
  >> across major service providers (no mom and pop shops in the middle or
  >> at the end points), how would you suggest I deal with that?
  >
  >Uh... Ignore it?
  >10% packet loss is quite within the normal range of parameters for a
  >packet switching network such as the Internet. If you want 0% packet loss,
  >you can lease your own private point-to-point lines.

  Uh... Michael, when we were running the NSFNET, as Hans-Werner and
  many readers of this list are well aware, we did _not_ accept 10% packet
  loss on any link or across the network. These problems stayed with
  the NSFNET NOC until resolution by the provider, MCI. We only considered
  -0%- loss to be acceptable.

  Sorry if others also responded to this,

  Steve Richardson/Merit