Some questions to think about in the wake of the Netflix/Comcast deal to pay for more direct connectivity to Comcast’s network.
Does this help or hurt Comcast’s effort to buy TW Cable? I’ve seen some suggest that it helps by removing the concern that Comcast might block traffic like Netflix. But it’s also a warning that the more marketshare a last mile provider has, the more leverage they might have to arm-twist other content providers into paying them for peering. I think the jury is out on how this affects cable consolidation.
Will Netflix have to make paid peering deals with all big providers over the next months and years? I’m guessing a few more at the least. Certainly Verizon isn’t likely to be backing down over at their peering points with Cogent anytime soon…
How big will a content provider’s or CDN’s traffic have to be before last mile providers try to make this sort of thing happen? And the followup, what about international PTTs, them too? No data yet on either front…
Does it have to be video traffic to trigger things, or are there other types of traffic that could get singled out as overflowing the poor, desperately expensive backbone peering connections? I think for now it’s probably video only, being asymmetric and easily identified. But if telepresence ever really takes off, or the next big thing emerges…
How big does a last mile provider have to be to be able to arm-twist a content provider? Is this something that is the privilege of only the very largest mile providers? Leverage…
Is this a market-based rate that we shouldn’t be concerned about? I think that whether the rate is or isn’t what they would pay to deliver the traffic another way isn’t the point. Comcast is a last mile provider there is not a market involved. Without an alternative, the price is determined by what they can get without being hammered in public or by the FCC.
Is this really new today? I don’t think so, the last mile was a natural bottleneck before and still is. Providers who own a big piece of it have been steadily getting more aggressive, and this isn’t the first, second, or even third salvo. It’s just the one that’s gotten noticed during this network-neutrality intermission. Leverage and power games have been a part of the peering/transit world since forever.
If you haven't already, please take our Reader Survey! Just 3 questions to help us better understand who is reading Telecom Ramblings so we can serve you better!
Categories: Cable · Internet Traffic
Join Our LinkedIn Group
In what other industry does a company spend billions to build , upgrade and maintain a service delivery platform in this case a network that reaches millions of homes and then be mandated to allow use by other company’s. VZ FIOS buildout cost them billions, Comcast upgrades cost them billions so we are sayingis thank You for beefing up platform so Netflix and others can use your network. Very tricky question and paid peering seems to be acceptable tradeoff if that cost in controlled and VZ is not allowed to gauge people like Netflix.
As a last mile provider, that’s how the Internet works.
Exactly. Firstly, it’s not like the Comcasts of this world don’t get govt assistance. Secondly, they should charge their customers the appropriate rate for the amount of traffic they consume.
What would you define as government assistance? Access to public rights of way is hardly a handout, it’s a business relationship between municipalities and service providers.
Also, it sounds like you’re arguing for metered service, is that right? (Because America LOVES pay-as-you-go?)
Besides, doesn’t charging Netflix amount to charging based on consumption of a shared scarce resource? Or should Comcast charge ALL customers more so they can reliably deliver video to customers they share with Netflix?
Is Netflix also wrong for charging ISPs for shifting content to caching servers at the edge of their networks to relieve pressure on their transit links?
Netflix isn’t charging any networks for anything.
I stand corrected. I was under the mistaken impression that their Open Connect CDN had a fee structure associated.
Whatever the case, it’s obvious that “how the Internet works” is not a static definition. Just as peering agreements have evolved over time, Big Content and Big Access players are going to negotiate relationships that reflect what’s at stake for each side.
“mandated to allow use by other company’s”
Netflix isn’t using Comcast’s network.
Comcast customers are using Comcast’s network.
Comcast is threatening to break that retail broadband product to extort money from companies whose business model depends on the Internet being the Internet and not some balkanized mess.
Comcast is charging twice for something the broadband customer is already paying for.
“In what other industry does a company spend billions to build , upgrade and maintain a service delivery platform in this case a network that reaches millions of homes and then be mandated to allow use by other company’s”
All of them. Power Companies, Water Companies all regulated. Most have pricing dictated at the state or local level.
Cable TV rates are regulated by local franchise authorities, and services are required by federal law to serve community needs and interests, none of which precludes any particular method of apportioning costs between subscribers and content providers who both benefit from and rely upon scalable access infrastructure.
Verizon has 9 million internet subscribers; 6 million of those on Fios. CMCSA has 21. Why does VZ warrant a paid peering arrangment like CMCSA?
Customer count has nothing to do with it. Traffic ratios do.
“Traffic ratios do.”
Says who? There seems to be different opinions out there when it comes to networks and settlement-free peering. Some require ratios, some do not.
Wouldn’t it be wise to state this is a matter of differing business models and not “how the Internet works”?
If this were truly “how the Internet works”, I’d expect this behavior from ALL transit networks, and not just a portion.
I’d imagine that there are very few to no settlement free connections between major players that don’t have some sort of traffic ratio condition.
I would imagine that most if not all of the other transit networks don’t have customers moving thousands of gigabits/s in one direction only or have sufficient other volume to make up for it.
I’ve seen transit carriers that move significant traffic have no problem establishing settlement-free peering without traffic ratio requirements. These are networks utilizing 100G, so I’d assume they move quite a bit of traffic on their backbone.
What you are implying is that the larger you get, the more you should be paid as an ISP, because you now have more leverage.
Sounds rather counter-productive for an industry to label that as “how the Internet works”.
Revenue and costs should be relative, not the opposite.
I don’t care if you move 10 megabit or 10 petabit, if it is within an agreed traffic ratio, it should be free.
If it isn’t, you’re a customer and you should pay for it.
Again, the traffic ratio requirement is a business decision from how it’s perceived. I don’t see the justification simply by “how the Internet works”.
Either call it a business decision or justify the statement “how the Internet works”. When I see many other networks (both large and small) follow a different business model, it’s not longer an industry standard.
Nothing wrong with having a specific business model. I do, however, see a problem with labeling an industry as a whole with behavior that is not consistent. When you label it as an industry standard when it clearly is not, it sounds more like dodging the fact that it’s a business decision (and there probably is a reason for that unfortunately).
If it exceeds that ratio, then you’re no longer a peer. Merriam-Webster definition for peer, “one that is of equal standing with another” Equal (or some percentage of tolerance. If you’re not exchanging equal traffic, then you’re not a peer.
Well, and also I guess if it bears you a cost now and the other side offers it for free, take it (which now makes me non-hypocritical).