The MPLS WG Archive

[Kireeti Kompella <kireeti@juniper.net>]

On Mon, 25 Apr 2005, George Swallow wrote:

> > This is one of the best written I-Ds for a long time!

Thanks.

> > Section 7
> > There are several specific codepoints defined in the draft and
> > marked as "to be defined by IANA". These are specific values
> > within the new name spaces defined in section 7. I think it would
> > help IANA if you listed all of the defined values in a central
> > place (perhaps with references to the defining sections).
>
> Good idea.  Will include.

Concur.  Should also make sure that there are allocation policies for
all the code point spaces (e.g., fields in Downstream Mapping TLV).

> > Appendix A.
> > I'm sure this one has been out and round the block several times.
> > Notwithstanding that the appendix is non-normative, why is it included? I
> > thought we were doing no new work (apart from drafts listed in RFC3468)
> > related to CR-LDP.
>
> Is there anyone who objects to removing Appendix A?

Not me.

> > Section 3.5
> > The Error Code TLV
> > I think no-one came up with a use for this. Since we try not to define
> > things unless we have an actual use, perhaps this TLV should be retired.
>
> I'll leave it to Kireeti to defend as I also suggested taking it out.

Take it out.

> > And a silly question.
> >
> > Suppose I have an RSVP-TE LSP that starts on an IPv6 LSR and ends on an
> > IPv4 LSR (or vice versa).
> > So I have an IPv6 Sender Template and IPv4 Session (or vice versa).
> > How do I encode the Target FEC Stack?
> > Or perhaps this is just an unrealistic deployment?
>
> Hmmm.  Good luck getting a notify message back!  Not something I was
> planning to support.

When there is an actual deployment of this, we'll re-evaluate whether
we need N*M RSVP-TE TLVs (N sender templates * M sessions) or N+M TLVs
that the ping-er can mix-and-match.  The two TLVs for IPv4+IPv4 and
IPv6+IPv6 should cover 99.314% of the cases, so we'll leave those
unperturbed.

We can re-evaluate this sooner, when we get to the point where we're
discussing the nuts-and-bolts of the p2mp RSVP-TE ping TLVs.

Kireeti.
-------

_______________________________________________
mpls mailing list
mpls@lists.ietf.org
https://www1.ietf.org/mailman/listinfo/mpls