UUNET Pulling Peering - Core Issues

Hong Chen hchen at aimnet.net
Sun May 4 01:19:41 UTC 1997



I think that what UUNet just announced has far more impact than
just forcing smaller ISPs to pay for the peering. 

Essentially, the umlimited flat rate service offered by smaller
ISPs will be gone, and replaced by metered services. That is, 
you pay what you use. 

Because of UUnet and other large ISPs' demand for peering ($60K/month
for DS3?), the operating cost to regional ISPs will go up. The 
regional ISPs will have to raise their fee. 

There are other problems that will come soon after UUNet's 
announcement. That is the issue of 'who should pay for the 
international lines'. International lines are quite expensive
and I remember that Tokyo Internet just paid US $20M for a 
4 year DS3 contract from Japan to US. 

According to many major ISPs overseas, today's US Internet Centric
status will not last long when Internet grows tremendously overseas, and 
some day, US national ISPs either have to build their own International
links, or they have to share the international link cost with 
the oversea providers, just like today's telecom model.

I can even see that global ISPs such as UUnet, MCI, etc one day
will charge large national ISPs in the US for peering fees as
UUNet and MCI think that they have paid the expensive international
links. 

What does this mean? It means that peering fee will likely go up 
again (bandwith also grows), and cost to the regional ISPs can be 
even higher. The regional ISPs will have to ask themselves 'does it
make sense to charge $19.95 for a customer while this customer
is sending unlimited number of emails worldwide, and making 
unlimited internet phone calls worldwide?' To cover the cost, 
should the regional ISPs raise the fee to $30/month for unlimited?
Maybe a tired fee structure will make sense, like what Netcom is
doing today.

The question begs again: Does the smaller ISP or regional ISPs 
have the necessary software/etc to bill for the metering or 
tiered services ....

Sounds like that a conference can be held for the afermath of 
'ending of free peering' 


Hong Chen
hchen at aimquest.com

On Sat, 3 May 1997, Charles R. Hoynowski wrote:

> 
> 
> Uh, excuse me?  I have no right to count on complete Internet connectivity
> from my ISP?  This has some far reaching implications.
> 
> What consideration is being given to arguably the most impacted parties in
> this new peering (or not peering) policy?  The business customers, selling 
> services via the Internet.  Consider, company "A" on Alternet is contracted 
> to provide data on demand to support services running on company "B", who 
> is on another ISP.  Without warning or notice to its customers (which may be 
> the most disturbing point) UUNET makes a policy decision to stop peering 
> with the other ISP.  Company "B" goes out of business.  "B" sues "A" for 
> breach of contract, "A" sues UUNET, etc.
> 
> Should ISPs be allowed to set peering policy for Internet?  Should they have
> the right to define what the Internet is?
> 
> Which points out how the Internet really has become a common carrier/utility
> company concern.  PacBell can't suddenly change your phone service because
> they decide to.  There's public review and approval steps they have to
> follow first.  And there's good reason for that.  The public relies on the
> phone system for business transaction and private use.  The system must be
> be maintained in a consistent manner to allow that to occur.  I don't need
> to point out the similar common carrier roles ISPs and the Internet as a
> whole have assumed.  
> 
> The issue has brought to light the fact than players in the ISP game want 
> to assume the role, and make profits based on, of providing common utility 
> services, without assuming the responsibility and accountability structures 
> other public service providers must assume.
> 
> I believe this is the core issue at hand regarding UUNET's peering policy
> changes and needs to be address at that level.
> 
> 
>     _/_/_/_/ _/_/_/_/ _/_/     _/  _/  Charles R. Hoynowski, Site Admin
>    _/         _/_/   _/ _/    _/ _/    Etak, Inc., 1430 O'Brien Drive,
>   _/_/_/     _/_/   _/_/_/   _/_/      Menlo Park, CA 94025
>  _/         _/_/   _/   _/  _/  _/     Voice:415-617-4458 Fax:617-0161
> _/_/_/_/   _/_/   _/    _/ _/    _/    Email: charles at etak.com	      
> 
> 






More information about the NANOG mailing list