Re: QOS Author is Motorola, Chief Software Architect

[David Farber <dave () farber net>]

________________________________________
From: Patrick W. Gilmore [patrick () ianai net]
Sent: Monday, June 23, 2008 10:20 AM
To: David Farber; Waclawsky John-A52165
Cc: Patrick W. Gilmore
Subject: Re: [IP] QOS  Author is Motorola, Chief Software Architect

This post shows what I have found to be a common problem in arguments
around QoS / NN / traffic on the Internet in general.

Please, please, please everyone, stop thinking "fiber" ==
"bandwidth".  It does not.


> From: Waclawsky John-A52165 [jgw () motorola com]
> Sent: Monday, June 23, 2008 1:08 AM

> Hi Dave, Some QoS perspectives that I have learned: First, the main
> problem. QoS really isn't needed when you have big pipes.  The
> Internet
> has plenty of capacity and most applications don't really need huge
> amounts of bandwidth to work well. Go to:
> http://www.networkworld.com/news/2007/021507-dont-expect-video.html
> and
> read the 5th paragraph. It begins with "In the long haul....".  So
> what
> is the average utilization of these big pipes. ...single digits???,
> and
> for lots of good reasons. I used to professionally work on this
> stuff in
> another life and what I remember about it below and I think most, if
> not
> all of it, still applies.

That article was written in February of 2007.  The Internet look very
little like it did over a year ago, so I am not sure I would quote
that article today.  But one thing on the 'Net is the same as 1 or
even 5 years ago: The amount of glass in the ground has little to do
with the bits transmitted on it.

Saying there is plenty of fiber is not a useful way to quantify
bandwidth.  That is like saying there is plenty of land on either side
of the freeway, so why is there traffic congestion?

Actually, it's worse than that.  Imagine your commute from home to
work every day takes you through stop-and-go traffic.  Now imagine
someone tells you there is a strip of freeway 10 miles long with no
development on either side which isn't even on your way to work.
Finally imagine he then claims this proves you should not have to stop
once from your driveway to your desk.

What exactly would you say to that person?


> The main problem I see is with the term itself: "QoS" is NOT about
> Quality and it is NOT about Service. It is about billing!

You are just as wrong as the people who say it is all about quality &
service.  There is a problem they are trying to solve.  Being good for-
profit companies, Internet backbones try to solve their problems _and_
make money at the same time.  In fact, making more money helps them
remove the root problem (CapEx/OpEx of running a network large enough
to support all bits).

It is not either/or.

[SNIP]

Interesting claims.  I note the complete lack of data to back up your
claim.

I really don't have the time to look at each of your points and
explain why they are right or wrong.  I just have a single question
for you:

Have you actually priced out the cost of additional DWDM chassises,
shelves, rackspace, power, cooling, then the additional router
chassises, ports, rackspace, power, cooling, the addtional engineers
to operate said equipment, etc. vs. turning on a feature extant in the
routers you own today and administering it in a global manner to
guarantee "QoS"?

I would bet significant cash you have not.


You don't like QoS, fine.  Not sure I like it either.  But your post
is not a useful rebuttal against the realities facing network providers.

To be clear, of COURSE some providers are claiming things which are
obviously silly, or creating the very problems they claim are
insurmountable.  However, there are lots of problems which are hard
(or at least require tons of money) to solve.  Focusing on the former
case does not make the latter case go away.  And unless you can
address both cases, your argument is not useful.

I haven't found a way to address both cases.  Hopefully some smart
person will.  And soon.  (But I'm not holding my breath.)

--
TTFN,
patrick




-------------------------------------------
Archives: http://www.listbox.com/member/archive/247/=now
RSS Feed: http://www.listbox.com/member/archive/rss/247/
Powered by Listbox: http://www.listbox.com