Re: [CCAMP] Objective function draft

["Adrian Farrel" <[e-mail]>]

<individual contributor mode>
 
I'm hoping that you have looked at RFC 5623 that describes some approaches to
inter-layer networking using PCE.
 
But it seems that perhaps you are more interested in the per-domain PCE
mechanism (RFC 5152) and that you have observed that the OF that would be used
in each domain to select the path is not currently available because it is
missing from RSVP-TE. That seems to me to be a perfectly reasonable thing to add
to RSVP-TE and I think it enables and supports the work of the PEC working group
without any conflict, but I would hope that you share definitions of OFs with
the PCE WG.
 
And one last point, when you say
 
> even if there is no PCE
 
you misunderstand the PCE architecture. If you are computing a path (in an LSR)
you are de facto providing PCE functionality. There is no need in the
architecture (and thus the implementation) to expose an external (PCEP)
interface in order to be logically providing PCE function, and this is *exactly*
why you want access to the OF.
 
However, you would perhaps do better to absorb this ambivalence and note that
you need the OF and Metric whether or not the PCE is external.
 
All of which you may take as general support for your idea, but a feeling that
the I-D is not happily written and deviating too much from shared context with
PCE.
 
Cheers,
Adrian
 
From: [e-mail] [mailto:[e-mail]] On Behalf Of George
Swallow (swallow)
Sent: 11 September 2012 20:29
To: Julien Meuric
Cc: [e-mail]
Subject: [CCAMP] Objective function draft
 
Julien -
 
Reading the CCAMP notes (which capture little of the actual discussion) I see
that there may have been a perception in the room that PCE functionality at the
UNI-N was assumed (actual or proxy).
 
This is not the case.  The reason for our draft is that with the UNI, much of
the functionality that resides at the headend is moved to the UNI-N.  So the
UNI-C needs a way to express an objective function even if there is no PCE.  
 
Operationally it seems burdensome to require a PCEP at the UNI-C and a PCEP at
the UNI-N, when all that is being done is enabling the UNI-N to perform what the
client would do if it were connected to the network via a normal link.
 
Do you still object to the draft?  
 
Thanks,
 
.George