On Aug 15, 2013, at 00:19 , Sean Donelan <sean@donelan.com> wrote:
On Wed, 14 Aug 2013, Patrick W. Gilmore wrote:
It is actually even harder than the above illustrates. Most people define "Mbps on the Internet" as inter-AS bits. But then what about Akamai AANP nodes, Google GGC nodes, Netflix Open Connect nodes, etc.? They are all inside the AS. Given that Akamai claims to be 20% of all broadband traffic, Google is on the same order, and NF claims to be 30% of US peak-evening traffic, it seems like it would be foolish to ignore this traffic.
I could go on, but you get the point. Definitions are a bitch.
Some of that may help explain why the Internet traffic estimates seem to be too high or too low since about 2007. The primary data sources for the Internet traffic estimates seem to be mostly Internet backbones and Internet exchange points.
I hadn't been paying attention until I looked at a bunch of companies' investor filings this week because the size of the Internet was in the news. If you add up the percentages that companies are telling investors and policy makers, you end up with more than 100%. Most of the companies' investor reports don't explain % of what. But the few that do, end up pointing back to the same traffic forecast reports. That doesn't even get to the "long tail" of small providers that don't report anything.
Either there is a lot of traffic missing, or market concentration is much greater than assumed.
I am not at all surprised the sum of percentages is > 100. User on Joe's-DSL-and-Bait store sends a packet up through Mary's-backbone-and-coffee shop to Bill's-other-transit-and-sandwich cart which finally lands on Comcast. (Didn't see that coming, did you? :) All four networks are going to claim that packet, but a true accounting of "petabytes downloaded per day" will only count it once. -- TTFN, patrick