Re: [RTG-DIR] RtgDir review: draft-ietf-idr-bgp-ls-segment-routing-msd-16.txt

[Alvaro Retana <[e-mail]>]

Mach:

Hi!

You have a good point, the information can come from other sources.
Besides Direct or Static, BGP is another defined source.

So far, all the BGP-LS RFCs that have been published assume an IGP origin.
Note that this document originally talked about carrying data originated by
BGP — but it would have been the first published document to do so…and
defining the general operation of a BGP source didn’t seem to best fit in
an extension document.  So it was agreed to document that in its own
document [1].

Your point about processing of BGP-LS, in general, at the consumer is also
a good one…and one that caused a lot go discussion related
to draft-ietf-idr-bgp-ls-segment-routing-ext (IIRC).  The result being
rfc7752bis.


Thanks!

Alvaro.


[1] https://mailarchive.ietf.org/arch/msg/idr/uiTEbxlhsfGX_A5D7RQRrrRuA-M/

On April 17, 2020 at 11:30:11 PM, Mach Chen ([e-mail]) wrote:

Technically, according to Table 2 of RFC 7752, the source of BGP-LS
information can be from IGP, Direct or Static configuration. And the rules
in question are not specified in the IGP RFCs, I personally think they are
essential. It’s pity that we did not catch it when progressing those RFCs.
I do agree this document may not be the perfect place to specify  those
rules, but with the current situation, to write down something here is
better than do nothing.