Having worked for an ISP I'd like to add some points. (Sorry if they have been covered already!)
1. Netflix is a big company, they make money from their customers (the service isn't free so I'll assume there is profit). Therefore if the traffic associated with this service is so high that it significantly affects the ISP's then ISP should get a cut of this profit to improve their capacity so that Netflix can still deliver their service to customers. If not then price of the internet connection will need to rise as the usage does.
2. ISP's are always struggling with costs of bandwidth. Peering costs are not really the issue, the costs for the access network are, i.e. BTw.
3. With the current model it is hard to build a CDN that works efficiently. The CDN needs to be in the access network, as close to the customer as possible to save costs. More like a web cache in a corporate company. For residential connections this needs to be in the exchange. Therefore Netflix should be installing servers in each exchange and offer direct peering to BTw and LLU operators from there, not just Telehouse etc.
4. Innovation isn't restricted by this issue as many think it is. Innovation whatever it is won't use much bandwidth in comparison until such time that it become the norm.
5. It shocks and annoys me that Which? are now campaigning for guaranteed speed on residential connections. They seem to completely forget that residential connections are only affordable because they are contended, best effort and have no SLA. If a residential customer needs a connection with guarantees then they should buy the correct product, a leased line.



Pages in this thread:
Print Thread
deleted