NNSquad - Network Neutrality Squad
[ NNSquad ] Re: Irish Times: "A modest proposal on internet neutrality"
" Applying section 106 of the NPRM to business models for inter-provider
connection ("paid peering") is a common misunderstanding; it was
actually dismissed as FUD by the FCC folk attending the Austin NANOG
conference:"
When did I ever claim 106 applied to inter-provider connections? Paid
peering is a product being sold to *content/service/application* providers
and it certainly qualifies as "enhanced" or "prioritized". If you ask the
FCC the wrong question whether it applies to inter-provider connections, of
course they're going to say that's nonsense. But that does not address the
problem with NPRM section 106 and I have yet to see the FCC specifically
clarify this section.
" Telia defines varying service classes for *on-net* traffic, dictating
how packets get queued/dropped in the event of backbone congestion.
This does not cover inter-provider connectivity."
Again, stop interjecting "inter-provider connectivity" into the discussion.
That isn't what is being banned by Net Neutrality and I never claimed it was
being banned. What I specifically mentioned was voluntary ISP to website
(service/content/application) agreements that are being threatened by Net
Neutrality which is has been adopted by the FCC proposal and Rep. Markey's
proposals in Congress.
What I am concerned about is the arbitrary and capricious nature of Net
Neutrality that specifically targets ISP enabled router prioritization. It
ignores CDN enabled prioritization which is functionally similar to a higher
DiffServ priority class
http://www.digitalsociety.org/2010/08/data-shows-cdn-prioritization-more-har
mful-than-router-prioritization/).
It ignores the prioritization of a particular website or service *if* it's
the end user doing it or *if* it's the end user paying the ISP to do this.
So if an end-user wants to pay AT&T to prioritize traffic from Vonage or
Blizzard, that's OK. But if Blizzard or Vonage wanted to pay the ISP to
prioritize the exact same traffic, even with explicit authorization from the
end-user, and which is effectively an indirect payment from the end-user but
cheaper because of fewer transactional costs and higher scale, then that
must be prohibited. This makes absolutely no sense and it is purely a
special interest play in the guise of "public interest".
"This, too, was addressed in Austin, in the context of provider VoIP
and various business services focused on headend-based prioritization.
I recall a clarification that these practices aren't automatically
prohibited under the NPRM, though there would need to be specific
disclosures and "opting in"..."
Given your propensity to confuse the debate by conflating inter-provider
connectivity, I don't have a lot of confidence in your hearsay.
Furthermore, I'll believe this when I see actual language in the NPRM. As
it stands now, I'm not going to rely on some obscure meeting that isn't even
documented much less binding.
George Ou
-----Original Message-----
From: Adam Rothschild [mailto:asr+nnsquad@latency.net]
Sent: Saturday, August 14, 2010 12:37 PM
To: George Ou
Cc: 'Vint Cerf'; 'Lauren Weinstein'; nnsquad@nnsquad.org
Subject: Re: [ NNSquad ] Re: Irish Times: "A modest proposal on internet
neutrality"
(Disclaimer: I'm with a hosting/content provider, soon to be
announcing its position on this topic in a public arena. It's
similarly important for readers to note, for purposes of establishing
bias, that George makes his money lobbying for big telecom/cable...)
On 2010-08-14-03:42:20, George Ou <george_ou@lanarchitect.net> wrote:
> Furthermore, this is already an accepted practice on the Internet. ISPs
> like TeliaSonera already sell access to Blizzard with enhanced priority.
> Business connections routinely have enhanced priority. Global Crossing
> sells enhanced priority to business customers and they even extend that
> priority to partner networks in Asia and this has been happening for a
while
> now. Who is Google or anyone to say this is wrong?
Prioritization *between* providers isn't really something happening
deliberately. The instances of it I've seen usually involved one
peering party inadvertently setting or flattening/rewriting DSCP tags,
which itself would be an interesting course of study.
Telia defines varying service classes for *on-net* traffic, dictating
how packets get queued/dropped in the event of backbone congestion.
This does not cover inter-provider connectivity. We're also Telia
customers, and concluded that their usual "best effort" service,
combined with a healthy assortment of multi-homing and vendor/path
diversity, served our needs just fine. :-)
I'm not familiar with the GX product, though I'd imagine it's a
similar deal. They're not really well-peered in Asia, on an IP
level.
I'm not sure how this is revelant to a discussion of net neutrality...
> The FCC's NPRM proposal bans charges for "enhanced or prioritized" access
to
> content/application/service providers and that is a pretty broad paint
> brush. That potentially outlaws a number of beneficial models [...]
Applying section 106 of the NPRM to business models for inter-provider
connection ("paid peering") is a common misunderstanding; it was
actually dismissed as FUD by the FCC folk attending the Austin NANOG
conference:
http://www.nanog.org/meetings/nanog48/abstracts.php?pt=MTUyMyZuYW5vZzQ4&nm=n
anog48
> Furthermore, a ban on Paid Peering harms smaller websites that can't build
> their own infrastructure and negotiate free peering. Is it a coincidence
> that this harms Google's competitors?
Here in the US, the folk commonly purchasing "paid peering" services
are actually mid- to large-sized networks, by traffic volume, with
infrastructure to support this kind of a connectivity arrangement.
It's relatively easy to deduce who these purchasers are, looking at
route advertisements and community tagging -- I could put together a
list of examples if it's something of interest.
Absent VC backing, the "two guys in a garage" will usually purchase
some form of hosting service(s) (virtual hosting, CDN, VMs, leased
servers, collocation, ...) from the providers in the above category.
:-)
It's important to consider that many access providers aren't strangers
to peering with content in a *settlement-free* manner. What you have
here are two sides of the equation who genuinely care about cutting
costs and ensuring the best possible delivery for their customers, all
*without* the incentive of money passing hands, nor the threat of
government intervention. I'd go so far as to classify "paid peering"
as a relative rarity, and an idiot tax of sorts, though this depends
largely on a specific organization's reach and business objectives...
(This is an easy point to miss if one's "research" is comprised of
simply reading the recent Norton whitepapers.)
> Wait, I thought Google cared about the "two guys in a garage"? Oh
> wait, that was just lip service and Google actually doesn't care.
I'm not convinced that Google or Verizon "sold us out", or otherwise
acted against the best-interests of the Internet ecosystem as a whole.
What they did was huge: showed that two diametrically opposed giants
can sit down and work together towards a common framework, all absent
any government intervention. While some of the specific points are
enough to raise eyebrows when taken at face value, I'd like to think
that these will be refined over time, and that common sense and
self-regulation will ultimately prevail.
As further proof that Google "actually cares", one need only look at
their contributions towards IPv6 implementation. I was fortunate to
attend their annual "IPv6 Implementors" conference in Mountain View
this past June, keynoted by Dr. Cerf, and was hugely impressed by
their knowledge-sharing and general interest in making the Internet a
better place. These are people clearly talking the talk *and* walking
the walk.
> Lastly, Net Neutrality doesn't even allow for user-approved
prioritization.
> If a user explicitly gives an ISP permission to prioritize a particular
> website or a general class of applications, who are you or anyone else to
> say no?
This, too, was addressed in Austin, in the context of provider VoIP
and various business services focused on headend-based prioritization.
I recall a clarification that these practices aren't automatically
prohibited under the NPRM, though there would need to be specific
disclosures and "opting in"...
-a