[128495] in North American Network Operators' Group
Re: off-topic: historical query concerning the Internet bubble
daemon@ATHENA.MIT.EDU (Jeff Young)
Tue Aug 10 19:29:18 2010
From: Jeff Young <young@jsyoung.net>
In-Reply-To: <AANLkTi=MsYAfA5EttFjpo71Bq1fUB+RSyBRgD4HEeHca@mail.gmail.com>
Date: Wed, 11 Aug 2010 09:28:48 +1000
To: Kenny Sallee <kenny.sallee@gmail.com>
Cc: nanog@nanog.org, Andrew Odlyzko <odlyzko@umn.edu>
Errors-To: nanog-bounces+nanog.discuss=bloom-picayune.mit.edu@nanog.org
-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA256
At the time these statements were made it was possible to make =
reasonable=20
assumptions about the size of the Internet. As a Tier 1 knew how much =
traffic our=20
customer links generated by the size of the link. We knew exactly how =
much
traffic stayed within our backbones and how much traffic ended up in a=20=
peering arrangement. We knew with some precision just how much of the=20=
Tier 2 ISP market was connected to us and how much was connected to =
others,=20
and who the others were. I don't think the theory still holds but =
traffic on-net
versus off-net was a pretty good indication of market share.
Today's Internet handles much more traffic in-region and is bounded by=20=
phenomenon such as language barrier (although the amount of spam I get
in Chinese characters has increased recently, who let the barrier =
down?). =20
This phenomenon wasn't as prominent in '98-'01 and while I wouldn't say=20=
it's impossible I think you'd have to commission the folks at UCSD to =
get=20
anything that resembled a value for total Internet capacity today.
Doubling in 9-12 months was a reasonable figure back then. 100 days
might have been a short-term spike caused by a back-log of activations
(we sometimes stopped the machine while we made upgrades) but it=20
certainly was an anomaly. =20
jy
On 10/08/2010, at 9:01 AM, Kenny Sallee wrote:
> On Fri, Aug 6, 2010 at 2:52 PM, Jessica Yu <jyy_99@yahoo.com> wrote:
>>=20
>> I do not know if making such distinction would alter the conclusion =
of your
>> paper. But, to me, there is a difference between one to predict the =
growth
>> of
>> one particular network based on the stats collected than one to =
predict the
>> growth of the entire Internet with no solid data.
>> Thanks!--Jessica
>>=20
>=20
> Agree with Jessica: you can't say the 'Internet' doubles every x =
number of
> days/amount of time no matter what the number of days or amount of =
time is.
> The 'Internet' is a series of tubes...hahaha couldn't help it....As we =
all
> know the Internet is a bunch of providers plugged into each other. =
Provider
> A may see an 10x increase in traffic every month while provider B may =
not.
> For example, if Google makes a deal with Verizon only Verizon will see =
a
> huge increase in traffic internally and less externally (or vice =
versa).
> Until Google goes somewhere else! So the whole 'myth' of Internet =
doubling
> every 100 days to me is something someone (ODell it seems) made up to
> appease someone higher in the chain or a government committee that =
really
> doesn't get it. IE - it's marketing talk to quantify something. I =
guess if
> all the ISP's in the world provided a central repository bandwidth =
numbers
> they have on their backbone then you could make up some stats about =
Internet
> traffic as a whole. But without that - it just doesn't make much =
sense.
>=20
> Just my .02
> Kenny
>=20
-----BEGIN PGP SIGNATURE-----
Version: GnuPG/MacGPG2 v2.0.14 (Darwin)
iF4EAREIAAYFAkxh4LUACgkQxvthcni5E2+7AwD+Lx+Dm14XTn/qZpy2co3CrcI1
dzA9QycoM2VmMBjmfxwA/1LD7gqI3zd80VozkHMDbDIREDPxKBPPtMlb+7Tu/nPV
=3Dwt/O
-----END PGP SIGNATURE-----
