If my network has huge connectivity between end users and individual cable head-ends, but contended bandwidth between the head-end and major internet interconnection points, I absolutely want to give customers an incentive to use the cable head end to customer bandwidth, rather than Internet to customer bandwidth.
(As far as I can tell, net neutrality only matters in the context of video (or possibly bulk file sharing of video); virtually no other traffic on wired networks is meaningful
enough in scope, scale of bandwidth used by users, or performance/latency sensitive enough to be an issue. This is essentially a Comcast vs. YouTube/Netflix argument.)
Colocating equipment at head ends isn't a solution, because 1) not every provider will colocate, so if you can't preference colocated streams over non-colocated streams from other similar services, there's much more impact on other users
2) it might not be feasible to offer third-party colocation in every part of your network where there is bandwidth oversubscription.
> Colocating equipment at head ends isn't a solution, because 1) not every provider will colocate, so if you can't preference colocated streams over non-colocated streams from other similar services, there's much more impact on other users
I've read this 10 times but I can't follow. In particular, who is this "you" exactly? And what kind of impact?
2) it might not be feasible to offer third-party colocation in every part of your network where there is bandwidth oversubscription.
This is not a concern for net neutrality but rather the individual ISPs. No ISP is obliged to offer any kind of hosting whatsoever.
"You" being a hypothetical local network provider (potentially comcast, but more relevantly a small startup)
The impact being video streams from netflix/youtube killing your network. A better example is probably a rural WISP; huge bandwidth from tower to the 300-500 max users, but limited backhaul from the tower to your hundreds of miles away central location.
On 2), if you don't offer colocation to third parties, but only to your content partners, that works well IFF you can incent your end users to hit those servers preferentially. If that means offering iptv from them for $20/mo which doesn't come out of your bandwidth cap, and having a 200 GB/mo cap for other traffic, that is a win for everyone except netflix. If you aren't allowed to say "iptv from our servers is cap-free, everything else is capped", users hitting Netflix will continue to saturate the backhaul and no one will use the local media server iptv.
(I'm not necessarily saying net neutrality is a bad idea (or a good idea), but that there are specific user-beneficial cases which net neutrality hurts. I'm concerned that Comcast basically can rape everyone today with the laws that are in place, and is entirely likely to use net neutrality laws to do so as well -- they're a perfect example of an incumbent being so big that regulations just serve to entrench them. Competition is the answer, and net neutrality in some ways helps competition among pure-IP cloud-delivered services, but hurts competition among access providers.)
(As far as I can tell, net neutrality only matters in the context of video (or possibly bulk file sharing of video); virtually no other traffic on wired networks is meaningful enough in scope, scale of bandwidth used by users, or performance/latency sensitive enough to be an issue. This is essentially a Comcast vs. YouTube/Netflix argument.)
Colocating equipment at head ends isn't a solution, because 1) not every provider will colocate, so if you can't preference colocated streams over non-colocated streams from other similar services, there's much more impact on other users
2) it might not be feasible to offer third-party colocation in every part of your network where there is bandwidth oversubscription.