Netflix To Cogent To World
asr at latency.net
Thu Jul 24 20:46:30 UTC 2014
Not to single out Jason, who has demonstrated his worth as one of the “good guys” in the community time after time, however I and somewhat of a skeptic:
That Comcast is in a “pretty good spot” for capacity could be punctuated by any number of shifts in traffic, or new sites/services emerging as the next killer app. Where other access providers would increase capacity, Comcast would see money in its eyes, or cite such dated metrics as traffic ratios as a fairness metric, all the while playing the victim with the press.
I don’t think I’m overly alarmist in these views; one need only look to the Tata situation (congested for multiple years), which was a textbook case of poor execution and damage control by all involved, as a recent example. Fool me once...
On Jul 24, 2014, at 1:00 PM, Livingood, Jason <Jason_Livingood at cable.comcast.com> wrote:
> On 7/23/14, 1:18 PM, "Adam Rothschild" <asr at latency.net> wrote:
>> Comcast¹s position is that they could buy transit from some obscure
>> networks who don¹t really have a viable transit offering, such as DT and
>> China Telecom, and implement some convoluted load balancing mechanism to
>> scale up traffic.
>> (I believe this was in one of Jason Livingood¹s posts to
>> broadbandreports, unfortunately I don¹t have a citation handy.)
> I¹m pretty sure I didn¹t say specifically that DT and China Telecom were
> options. I probably pointed out the lack of delivery problems prior to
> using delivery partners like Cogent (such as via Akamai or Limelight) and
> that delivery alternatives existed. But that¹s in the past - we¹re in a
> pretty good spot w/Netflix traffic right now, though we continue to add
> capacity as you¹d expect.
More information about the NANOG