Other than misuse of the term "peering," it was an interesting article in the light of net neutrality decision.
Here are some posts on "peering" from another poster, ivanhoek
, who was in the industry from another thread about 6 months ago. Hope the explanation would provide some color and context to the article. It explains why Youtube and Netflix Open Connect CDN would be relatively unsuccessful. I am sure many of us remember his posts.
Originally Posted by ivanhoek
I have worked for major US ISP's and MSO's in the past, as well as Microsoft and DDOS mitigation companies , and I can say this.. ISP's agree to peer at the major exchanges when they believe that both partners in the peering agreement are more or less equal from a transit standpoint, that is, ISP A is likely to send as much traffic to ISP B, as ISP B is to send to ISP A. The peering relationship then makes sense, since it would not impact revenue (the transit fee revenue loss is offset by the savings of not paying out the same), and it simplified billing relationships.
Netflix isn't a transit carrier, so this arrangement doesn't work. Netflix will send the ISP a TON of traffic, but won't take any in return. Calling this peering is a severe misnomer and likely where all the major ISPs are balking.. There's zero gain for the ISP, and there is revenue loss from not charging transit for Netflix transit.
Actually, not unreasonable at all. Most, if not all of the ISP's who have joined open connect are smaller ISP's who buy transit from the larger tier1 carriers (consumers of transit), rather than sell transit. Most of those ISPs only sell access to subs and are in no position to charge for transit.
It's probably cheaper for the smaller ISP to get the traffic at the peering exchange directly from netflix than pay transit fees to their upstreams... remember, ISP's peer in tiers, and those smaller ones likely peer among themselves, but not with the big tier1's that sell them transit.
The bandwidth to the exchange is not free, so a small ISP also has the advantage of needing a smaller pipe, perhaps a single 10g wave. A big carrier would likely need to pay to augment their bandwidth at the exchange to absorb the presumably large amount of traffic they'd suck in from Netflix.
Also, these big companies have a LOT of projects competing for resources. A project that brings in no revenue, and doesn't align to the strategic direction has very little chance of getting started and resources allocated. There isn't a single MSO out there that has as its strategy commoditizing its bandwidth, and being what they call a "dumb pipe". They ALL want to deliver value added services and content. So most of the resources are going to projects that bring in revenue and/or align to the strategy.
This Open Connect has those two strikes (no revenue AND no strategic alignment) against it, when seen from the eyes of a major MSO.