Hi Igor,
On Oct 17, 2005, at 9:21 AM, Igor Bryskin wrote:
Adrian,
My personal preferrences:
PCE - Path Computation Entity - entity involved in PCE Architecture, a
common term for PCS and PCC
PCC - Path Computation Client - entity requesting path computation
PCS - Path Computation Server - entity that can satisfy path
computation
request
In your view PCE = PCS, and hence PCE Architecture is still mostly
about
servers. Yes, you did put a section about PCC, but PCC is still
kind of a
step-son compared to PCS. I beileve PCC is as important as PCS, and
PCE
architecture is about PCCs as about PCSs. Therefore, PCS is a
tighter and
less confusing term for the path computation server than PCE.
I fully agree with Adrian: there is no need to re-introduce the term
PCS here which by the way did generate a lot a confusion several
years ago.
Thanks.
JP.
Igor
----- Original Message -----
From: "Adrian Farrel" <[EMAIL PROTECTED]>
To: "Gray, Eric" <[EMAIL PROTECTED]>; "Lou Berger"
<[EMAIL PROTECTED]>
Cc: <[EMAIL PROTECTED]>; <[EMAIL PROTECTED]>
Sent: Sunday, October 16, 2005 6:22 PM
Subject: Re: [Pce] WG Last Call ended on draft-ietf-pce-
architecture-02.txt
Hello,
Nice to see some bandwidth being used.
"PCE Architecture" -- an architecture that includes a PCE.
"PCE" -- something that performs path computation.
"PCC" -- something that requests path computation.
In the Abstract...
This document specifies the architecture for a Path Computation
Element (PCE)-based model.
Is it really necessary to then state... "We will term this
architecture
the PCE Architecture" ?
I believe that at no other point in the architecture document is
there a
reference to "PCE" that means "PCE architecture" except it says
explicitly
"PCE architecture". I would be happy for you to point out places
where
this is a problem.
In other documents (such as draft-ietf-pce-comm-protocol-gen-
reqs-02.txt)
appear to make exactly the same distinction.
Thus, there is no use of the term "PCE" to mean an architecture.
It is
always used to mean a network element. There is no need to (re-)
introduce
the term "PCS" because it is isomorphic to "PCE" as currently
defined.
Thanks,
Adrian
----- Original Message -----
From: "Lou Berger" <[EMAIL PROTECTED]>
To: "Gray, Eric" <[EMAIL PROTECTED]>
Cc: <[EMAIL PROTECTED]>; <[EMAIL PROTECTED]>
Sent: Saturday, October 15, 2005 3:57 PM
Subject: RE: [Pce] WG Last Call ended on
draft-ietf-pce-architecture-02.txt
At 06:06 PM 10/14/2005, Gray, Eric wrote:
Lou,
What would be the point? Then it is merely a traffic
engineer application.
--
Eric
Eric,
The PCE Architecture doesn't limit PCE to multi-domain
networks. to quote:
"... Path computation in large, multi-domain, multi-region or
multi-layer networks is complex and may require special
computational components and cooperation between the different
network domains."
Lou
_______________________________________________
Pce mailing list
[email protected]
https://www1.ietf.org/mailman/listinfo/pce
_______________________________________________
Pce mailing list
[email protected]
https://www1.ietf.org/mailman/listinfo/pce
_______________________________________________
Pce mailing list
[email protected]
https://www1.ietf.org/mailman/listinfo/pce
_______________________________________________
Pce mailing list
[email protected]
https://www1.ietf.org/mailman/listinfo/pce
