[CCAMP] OIF draft on ason routing

["Ong, Lyndon" <[e-mail]>]

Hi Folks,
 
I wanted to give people a quick introductory email on the OIF-sponsored
draft on ason routing (draft-ong -gmpls-ason-routing-exper-01.txt) with
some background:
 
1.      The OIF recognizes that the IETF is the owner of standards for
GMPLS protocols, including especially OSPF.  It is not in OIF's scope to
define standards.  In fact, OIF is currently on a track to adopt a
subset of the OSPF routing extensions defined by IETF CCAMP in the
forthcoming Experimental RFC 5787 (aka
draft-ietf-ccamp-gmpls-ason-routing-ospf-09.txt, as the RFC has not yet
been published).
 
2.      That being said, we think there is a great opportunity for
synergy here if the IETF would consider adopting the formats defined in
the OIF draft as a testing-based update to the work in
draft-ietf-ccamp-gmpls-ason-routing-ospf-09.txt, as that draft suffers
from lack of implementation and testing (thus the Experimental status),
while the formats defined in the OIF draft have at least 8
implementations and the benefit of about 7 years of Interop testing.  We
believe this would accelerate the adoption and deployment of standards
for ASON routing using OSPF, and also save a good deal of effort and
resources on the part of OIF members to re-implement and re-test ASON
routing extensions. Also, given the long experience with running code,
it would be justified to move this work from Experimental status to
Standards Track.

3.      As explained in draft-ong -gmpls-ason-routing-exper-01.txt, we
think that the format modifications involved are quite small in the case
of the Local/Remote TE Router_IDs and the advertisement of local Address
Prefixes, basically in the one case the use of 2 vs. 1 combined sub-TLVs
and in the other case the use of address length rather than network mask
for the prefixing.  The issue of the ISCD for SONET/SDH is trickier, as
an alternative method is already defined in RFCs 4202 and 4203, however
the draft identifies concerns with the amount of overhead required using
RFCs 4202/3 and advertising multiple ISCDs for a fairly typical
SONET/SDH link of OC192 rate (which requires 5 ISCD sub-TLVs to be
advertised) and requests consideration of a supplementary sub-TLV that
would avoid needing to advertise multiple ISCDs.

4.      It has been suggested that OIF deliberately deviated from the
IETF work on ASON routing.  This is not the case.  Work on formats used
in OIF experimentation on ASON routing dates back to 11/2002, prior to
IETF CCAMP activities on ASON routing extensions, which were initially
submitted in June 2006.    

5.      One might ask why OIF did not submit input previously on this
issue.  I accept some of the fault for this, as I believed it was
implied to me that OIF input on formats was not especially welcome, and
that the fastest way to get standard OSPF extensions defined was to
leave the process entirely within IETF.   It should be noted that OIF
participants contributed to the work on ASON Routing requirements and
evaluation in RFCs 4258 and 4652 that preceded the work on OSPF
extensions.

We would welcome any comments on the draft, now or during its
presentation on Tuesday at CCAMP and are hopeful that this will be
received as constructive input.

Thanks,
Lyndon