RE: Accounting in peered CDNs: Initial thoughts ....

["Reinaldo Penno" <rpenno@nortelnetworks.com>]

you# -----Original Message-----
you# From: Hilarie Orman [mailto:HORMAN@novell.com]
you# Sent: Thursday, January 18, 2001 11:49 AM
you# To: Reinaldo Penno; cdn
you# Cc: Abhi.Deshmukh; ietf-openproxy
you# Subject: RE: Accounting in peered CDNs: Initial thoughts ....
you#
you#
you# I'd not mix the concept of information about traffic
you# requirements with
you# information about the *user*.

Well I give many types of information. Is is up to the device (surrogate or
traffic interception) to decide exactly what do with it (billing, acct,
traffic priotizations, etc).  Since it includes QoS information and
geographic location, it can help to enhance the request-routing mechanism
and allow CDNs to provide different types service (based on the information
provided)

And this is probably not a good
you# forum to discuss giving the user's name to traffic interception
you# devices.  If networks have been able to get my packets delivered
you# for the last 30 years without knowing my name, I fail to see why
you# they need it in the future.

Well, I do not know if you read the abstract but this if for billing and
acct purposes. Also, I would say that personalization of content is a good
thing. And I would also say that for 30 years you were able to get your
packets anyway so why the heck caches, diffserv, MPLS, etc  were proposed?

Also, this list/WG that you are part of, is trying to propose request
routing mecahnisms,etc. But since everybody here was able to get their
packets anyway in the last 30 years I wonder if they are wasting their
time...

you#
you# However, I can see this being used for user-centric services