Re: [Teas] use of the FEC term in draft-ietf-teas-lsp-diversity

Lou,

On 2015-04-17 02:01, Lou Berger wrote:
> Authors/WG,
>      I'm a bit uncomfortable with the document's use of FEC in the
> context if RSVP-TE.
Yes, I share that concern, the FEC terminology originated in LDP
context, but have since developed developed.

> RFC3209's usage of FEC is very loose and rfc4379
> only gives an indirect definition in one context.  I think that either
> the draft should avoid the term or it should point to a to be written
> formal definition of the term in the context of RSVP-TE (including in
> its GMPLS form).

I've been looking for a while for that defintion, and as far as I can
there is no RSVP-TE message or object that actually carry a FEC. It has
been said the the FEC definition is "implicit", but I don't think that
is a solution, rather it is part of the problem.

I think that "avoid" would be easiest, but if the authors go for
defining a FEC in RSVP-TE context, I think that the consensus call for
such a definiton need to involve teas and mpls, but I think ccamp and 
pce should also be notified for the such a consesus call.

/Loa
>
> Thoughts?
>
> Lou (with any / all hats)
>
>
> _______________________________________________
> Teas mailing list
> Teas@ietf.org
> https://www.ietf.org/mailman/listinfo/teas
>

-- 

Loa Andersson                        email: loa@mail01.huawei.com
Senior MPLS Expert                          loa@pi.nu
Huawei Technologies (consultant)     phone: +46 739 81 21 64