off-topic: historical query concerning the Internet bubble

Roland Perry lists at
Mon Aug 9 20:29:12 UTC 2010

In article <4C6030BF.1030306 at>, Nick Hilliard 
<nick at> writes
>On 09/08/2010 16:12, Christopher Morrow wrote:
>> I think, from another list about 2 yrs ago, the person responsible for
>> this data inside the company at the time (now not there) said someone
>> misinterpreted his stats/numbers...
>No doubt this is true.  And I note we haven't even started discussing 
>whether this "doubling every N time periods" refers to bit-rate, bytes 
>passed or daily max.

In the case of LINX traffic it was bits-per-second; any of: average, 
peak or total per day, because the shape of the daily and weekly curve 
didn't change much even over a period of years.

>I would have said that most networks during that period had occasional 
>burst growth rates of up to 100% within 100 days. The growth curve 
>second derivative is usually much bumpier than the first derivative. 
>So, regardless of source, the quotation is a truism, an urban myth and 
>ultimately means very little.

Growth at LINX was extremely steady (being an aggregate of over a 
hundred operators).
Roland Perry

More information about the NANOG mailing list