Re: A historical aside

From: Sean Doran (smd@ebone.net)
Date: Tue Dec 18 2001 - 13:53:30 EST


Fred -

| (although one could discuss MPLS COS; we're not developing it
| in a market vacuum).

No, indeed, the energy density in that parcitular product
space is very high indeed. But how can something be so
hot and yet still suck so much? A question for Guth[1].

| There are several ways to look at Henning's graph. One is to say "they
| obviously should purchase more bandwidth". I don't think you will find any
| argument that the money would be well spent. They may not have the money,
| though

In which case they don't have money to pay me to twiddle the
knobs to rearrange the chronic queues forming at the bottleneck
from me to them...

| or they may deem a link
| that is 100% utilized much of the day to be acceptable.

A 100% utilization with >> 0 packet loss & queueing delay is not
only acceptable, it's surely *desirable*.

An important observation:

Load != congestion

        Sean.

[1] http://nedwww.ipac.caltech.edu/level5/Guth/Guth_contents.html
    and in particular
    http://nedwww.ipac.caltech.edu/level5/Guth/Guth3.html
    which deals with "false vacuum".

    I think MPLS marketing is in "false non-vacuum", which appears
    to have an incredibly high relative energy density, but actually
    contains nothing material. But then I am bigoted.



This archive was generated by hypermail 2b29 : Mon Aug 04 2003 - 04:10:03 EDT