[ NNSquad ] Re: [OT?] NN definition(s?)

["Fred Reimer" <freimer@ctiusa.com>]

I don't believe you will find anyone but Brett and other ISP's in
agreement with your definition or goals.  I understand very well
the technical challenges that ISP's face with the introduction of
bit torrent like applications, but am unsympathetic with their
efforts so far to deal with it.  I'm not saying, and am in no
position to say, that ISP's are not welcome here, but rather that
they should not be surprised if they do not receive a lot of
sympathy or agreement with their views.  After all, the whole
group and debate is a result of the actions that ISP's have
taken.  It's only logical that there is disagreement with those
actions, for without disagreement there would not be a group!  I
don't believe politely explaining the actions is going to change
the minds of end-users, particularly those with a technical clue,
as to what is right and what is wrong.

Forging packets and sending RST's is bad.  The use of a new
technology such as P2P in and of itself is not.  I had to FTP
Fedora 8 from usf.edu just yesterday because I'm on ComCast and I
got a whopping 4K download speed with bittorrent.  I've never
used it for any illegal activity, and actually just installed it
specifically do download Fedora.  Does this tick me off?  You
bet.  Do the complaints from some ISP's alleviate the issue, or
relieve my distress, heck no.  So what did I do?  I nailed up a
bunch of FTP sessions to download not only the Fedora 8 install
DVD, but also the recovery CD and all of the "spins" that I could
get a hand on.  Instead of allowing the torrent, for which I
would have been happy with a 200K or so download speed, ComCast
blocked it, and resulted in 4 consecutive FTP's at 200K each.
You go ComCast!  Great traffic engineering there!

So, let's lay off the long unnecessary discussion of what net
neutrality is and whether it is good or not.  I agree with the
purpose of this group, to identify and document any preferential
or obstructive behavior of ISP's.  That's it.  Nothing more.  Let
other groups and/or legislators deal with what is done with the
information.  If you don't agree with that mission statement then
you may not find yourself welcome here, or find it a productive
use of your time.  Again, I'm not saying you are not welcome or
you should not participate if you are an ISP, just don't expect
agreement any sympathy from the masses of end-users that are
detrimentally affected by your actions.

Now with that said, I wish there was Another group to deal with
just what some appear to want to discuss.  That is, what SHOULD
ISP's do about P2P traffic. That would be an interesting
discussion that I may find technically challenging and may
participate in.  I don't think a group such as one named
"NNSquad" is appropriate for those types of questions or
discussions.  I do think such discussions are needed, and that
ISP's can do well to listen to input from the general public,
some of which do have a clue and would be willing to help devise
the technical response to such traffic.  Some of this has already
been discussed here and on other groups such as nanog.  That is
primarily QoS mechanisms and/or different pricing models.  But
again, such discussions are completely inappropriate for this
list, IMO.  You may agree or not, but there, I said it.

Thanks,


Fred Reimer, CISSP, CCNP, CQS-VPN, CQS-ISS
Senior Network Engineer
Coleman Technologies, Inc.
954-298-1697



-----Original Message-----
From: nnsquad-bounces+freimer=ctiusa.com@nnsquad.org
[mailto:nnsquad-bounces+freimer=ctiusa.com@nnsquad.org] On Behalf
Of Richard Bennett
Sent: Sunday, November 11, 2007 12:13 AM
To: The Anarcat
Cc: Brett Glass; David P. Reed; nnsquad@nnsquad.org; Robb
Topolski
Subject: [ NNSquad ] Re: [OT?] NN definition(s?)

On Sat, 2007-11-10 at 19:10 -0500, The Anarcat wrote:
> In a lot of anti-NN litterature I've seen, the "lack of a
proper
> definition" has always been a key attack point. For me, it's a
moot
> point: NN proponents have a fairly consistent definition, in my
opinion.
> 
>
http://en.wikipedia.org/wiki/Network_Neutrality#Definitions_of_Ne
twork_Neutrality
> 
> Can we move on now?

The Wikipedia section you cite gives three different classes of
definition for NN, which I recall distinctly because I wrote it.
Three
major classes is hardly "a fairly consistent definition," it's a
rat's
nest for policy makers.

So yes, the first order of business for any group that aims to
protect a
principle is to define the principle. If that can't be done, then
it may
as well pull up stakes and go home.

I had hoped that this group would bring some light to a debate
that has
so far been almost exclusively heat. So far, I don't see that
happening,
except insofar as Brett has tried (valiantly, in my estimation)
to
remind the group that ISPs aren't necessarily sitting on oceans
of free
bandwidth that they hoard in order to keep the troops in Iraq,
the
oceans overheating, and the dictators in power, etc.

If we understand neutrality as meaning the ability of the
typical,
non-abusive network user to access the sites and applications he
wants
within the constraints of the laws of physics and the state of
network
engineering, then we have to accept the fact that the
infringement of
this ability doesn't come exclusively from ISPs or monopolistic
phone
companies. Bandwidth hogs, spammers, worm artists, and other
sources of
noise also interfere with network access in this world of shared
facilities and pooled bandwidth.

Here's one example: the Azureus implementation of BitTorrent
allows the
user to set a global TOS/DSCP label for all his transmit traffic.
By
default, this value is set to a value that uses the default value
set by
the OS, which in most cases is "Best Effort" in WiFi terms. 

I would contend that BitTorrent is the best example of
"Background" or
"High Througput" traffic ever devised, so its default TOS is too
high.
And because it's too high, it interferes with the user's own
VoIP,
video-streaming, and web browsing traffic unless the user has
Phil
Karns's sophistication and equipment (unlikely, of course.)

If others agree that this is an abuse of net neutrality, perhaps
we can
proceed to publicize it and have it corrected after a period of
proper
flogging and public shaming.

Just a thought,

RB

Attachment: smime.p7s
Description: S/MIME cryptographic signature