I DID indeed mean one opes intermediary talking to another. I did NOT mean an
opes intermediary talking to a call-out sever. Based on Markus's message, I am
now clear in my mind that the problem that I feared does NOT exist.
"Menon, Rama R" wrote:
By stating "OPES 1 talking to OPES 2", what exactly do you mean?
(1) an OPES Intermediary talking to one of the OPES callout servers?
Or is it
(2) one OPES intermediary talking to another OPES intermediary?
I don't think the latter case #(2) is what you meant and I believe it is for
case #(2) that Jayanth originally asked the question on e-e discovery.
From: Markus Hofmann [mailto:hofmann(_at_)bell-labs(_dot_)com]
Sent: Wednesday, August 15, 2001 1:01 PM
Cc: Hilarie Orman; ietf-openproxy(_at_)imc(_dot_)org
Subject: Re: Edge to edge discovery
If OPES1 knows it is talking to OPES2, and *not* the ultimate
client of the session, then there is * no* problem.
This indeed is my understanding of the scenario - the session from
OPES1 would be properly terminated at OPES2, i.e. OPES1 is aware that
it's talking to OPES2.
However, we recently had discussions whether OPES devices should be in
the data path for streaming media, or whether it would be preferable
to just have them in the control path in this case (see discussion at
the OPES Workshop in June this year, minutes are at
Jayanth P. Mysore
Networks and Infrastructure Research Laboratory,
Phone : (847) 576-8561