[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: I-D ACTION:draft-ietf-ccamp-lsp-stitching-01.txt



Hi,

[i-d-announce@ietf.org and Internet-Drafts@ietf.org omitted from the cc
list]

[SNIP]

> > [dp] i would simply recommend to state what an LSP segment is (as
stated,
> > this comparison has been used to show differences in terms of control
> > plane processing and keeping it at that level is sensible) and not
embark
> > this document into terms and complex comparisons that are at the end
of
> > no real help
> <AA>----> I tend to agree with this.

Yes. If there is a need to compare the two techniques or to do
applicbality then that is another I-D. I am not sure that there is a need.

I would like to see this I-D restricted, as Arthi said, to a description
of what Stitching is and how it works.

> > 2. Why are you saying that a TE Link based on S-LSP can be used for
> > exactly
> > one e2e LSP? Parallel S-LSPs could be advertised as a single TE link
> > (bundle, see above) and hence can accomadate several e2e LSPs.
> >
> > [dp] i think the purpose is to say that a "triggered" LSP segment can
be
> > used by a single end-to-end LSP compared to the situation occurring
with
> > FA, where the triggered FA-LSP can then carry multiple nested LSPs
> <AA>------> As I mentioned in the email to Igor, the sentence
immediately
> following this does talk about the bundling case. What is missing is
> probably an additional sentence which explicitly states that when such
> bundling is use, more that one e2e LSP can be admitted over the LSP
> segment. That can be clarified.

Say rather, "can be admitted over a TE link supported by LSP segments".
That is, a segment can only admit one LSP, but a bundle of segments
advertised as a TE link, can admit multiple LSPs.

A