> Based on industry norms with respect to peering and its
> experience with the Level 3 gambit, Netflix knew that
> transferring its traffic to settlement-free routes would
> result in network congestion by dramatically altering the
> ratio of traffic exchanged.
This is a poorly crafted hit piece. The "ratio" of traffic exchanged wouldn't have changed. At most it would merely have shifted origins--from whatever link Akamai and Comcast shared, to the Cogent-Comcast link. Comcast is principally an edge ISP to content consumers, so Netflix's change in providers meant nothing to them in terms of usage patterns. And because it's a good bet that the exchange facilities were the same in all scenarios, setting aside potential contractual issues it likely resulted in no change in network traffic whatsoever.
I have no idea if Netflix' CDN strategy was also a maneuver in the regulatory realm. I would hope they were smart enough, at least, to have considered the potential conflicts beforehand. OTOH, it would be extremely stupid and risky (and thus highly unlikely) for the change in providers to have been intended primarily as a regulatory gambit. For all we know it was Comcast who strategically saw an opportunity to test the regulatory waters.
What I do believe with some confidence given the context and errors is that the article is a poorly written hit piece, perhaps paid for directly or indirectly by Comcast as part of its marketing and regulatory strategy, if not actually written by Comcast.
> When it decided to send its traffic solely through
> settlement-free transit providers in 2012, Netflix knew
> full well that its decision to congest peering and degrade
> the quality of its service would affect some ISPs very
> differently than others. Netflix had also started its own
> CDN in 2012, and some ISPs — those who do not operate
> Internet backbone networks, and thus would otherwise have
> to pay transit fees to receive Netflix traffic — had
> already agreed to the unusual arrangement of directly
> interconnecting with Netflix’s CDN without charge. Those
> ISPs would not be adversely affected by Netflix’s hostage
> strategy, and Netflix knew it. Netflix intended to cast
> blame only on the ISPs that operate Internet backbones,
> and thus would not realize the same cost savings from
> Netflix’s CDN as the ISPs who do not operate backbone
> networks.
The article tries to make the point that Netflix was trying to shift the cost of paying for bandwidth from themselves to Comcast by rearchitecting their peering arrangements in such a way that they benefited from settlement-free peering arrangements. But they just admitted above that it's usually the edge ISP who pays for traffic.
It's obvious that the company rearchitecting their peering in an attempt to shift costs is Comcast. Comcast is (indisputably) an edge, almost entirely consumer-facing ISP. They can refer to parts of their network as a "backbone" however and whenever they like, but both Comcast and this article are using the term "backbone" simply as a way to try to equivocate their contractual relationships with classic transit providers. But as the article indirectly explains, this norm exists largely because for classic transit providers load generation is symmetric--transit providers with a good mix of customers are similarly situated in terms of economic position, and it's cheaper to minimize bookkeeping with settlement-free arrangements.
But calling part of Comcast's network a "backbone" doesn't magically make Comcast's ingress and egress traffic symmetric. Yet the argument in the article is predicated on people making that erroneous mental leap. And as the article helpfully points out, when usage is asymmetric it's the edge ISP (on both ends) who ends up paying the bill. Thus, following the logic of the article as well as real-world practice, using settlement-free peering would normally be to Comcast's _benefit_!
It also follows that because it was Comcast who rearchitected their network (if in name only) to shift costs, that it was Comcast maneuvering strategically in the regulatory space. That doesn't mean Netflix wasn't making regulatory moves, but if so I'd characterize that as Netflix calling Comcast's hand at a time & place of Netflix' choosing rather than waiting around for Comcast to show its hand. In any event, from a purely business standpoint, no way would the _potential_ benefits that might accrue to Netflix be sufficient to justify such a change in transit and CDN providers.
Plus, this all ignores the fact that Netflix made the same free direct CDN peering offer to Comcast it did to every other ISP, which would have cut Cogent and Level 3 out entirely. It's just that Comcast rejected that offer; it wanted to be _paid_ for the peering! And they also rejected alternative offers to add more settlement-free ports with Netflix's transit provider. And Comcast's logic was precisely the specious argument that Comcast, purely by dint of how they referred to themselves, was situated like Level 3 or Cogent--a middle-man transit provider--rather than like, well, Comcast. But that's patently preposterous.
Of course, Comcast should be free to require payment to peer directly with Netflix. And they can be free to reject offers of more settlement-free ports with transit provider...
4 comments
[ 3.1 ms ] story [ 21.3 ms ] threadI have no idea if Netflix' CDN strategy was also a maneuver in the regulatory realm. I would hope they were smart enough, at least, to have considered the potential conflicts beforehand. OTOH, it would be extremely stupid and risky (and thus highly unlikely) for the change in providers to have been intended primarily as a regulatory gambit. For all we know it was Comcast who strategically saw an opportunity to test the regulatory waters.
What I do believe with some confidence given the context and errors is that the article is a poorly written hit piece, perhaps paid for directly or indirectly by Comcast as part of its marketing and regulatory strategy, if not actually written by Comcast.
It's obvious that the company rearchitecting their peering in an attempt to shift costs is Comcast. Comcast is (indisputably) an edge, almost entirely consumer-facing ISP. They can refer to parts of their network as a "backbone" however and whenever they like, but both Comcast and this article are using the term "backbone" simply as a way to try to equivocate their contractual relationships with classic transit providers. But as the article indirectly explains, this norm exists largely because for classic transit providers load generation is symmetric--transit providers with a good mix of customers are similarly situated in terms of economic position, and it's cheaper to minimize bookkeeping with settlement-free arrangements.
But calling part of Comcast's network a "backbone" doesn't magically make Comcast's ingress and egress traffic symmetric. Yet the argument in the article is predicated on people making that erroneous mental leap. And as the article helpfully points out, when usage is asymmetric it's the edge ISP (on both ends) who ends up paying the bill. Thus, following the logic of the article as well as real-world practice, using settlement-free peering would normally be to Comcast's _benefit_!
It also follows that because it was Comcast who rearchitected their network (if in name only) to shift costs, that it was Comcast maneuvering strategically in the regulatory space. That doesn't mean Netflix wasn't making regulatory moves, but if so I'd characterize that as Netflix calling Comcast's hand at a time & place of Netflix' choosing rather than waiting around for Comcast to show its hand. In any event, from a purely business standpoint, no way would the _potential_ benefits that might accrue to Netflix be sufficient to justify such a change in transit and CDN providers.
Plus, this all ignores the fact that Netflix made the same free direct CDN peering offer to Comcast it did to every other ISP, which would have cut Cogent and Level 3 out entirely. It's just that Comcast rejected that offer; it wanted to be _paid_ for the peering! And they also rejected alternative offers to add more settlement-free ports with Netflix's transit provider. And Comcast's logic was precisely the specious argument that Comcast, purely by dint of how they referred to themselves, was situated like Level 3 or Cogent--a middle-man transit provider--rather than like, well, Comcast. But that's patently preposterous.
Of course, Comcast should be free to require payment to peer directly with Netflix. And they can be free to reject offers of more settlement-free ports with transit provider...