Re: [OSPF] AD review of draft-ietf-ospf-sbfd-disciminator-03

[Alia Atlas <[e-mail]>]

Sounds great.   Do submit it.    Version numbers are cheap.

Thanks,
Alia
On Apr 13, 2016 12:39 PM, "Carlos Pignataro (cpignata)" <[e-mail]>
wrote:

> Hi Alia,
>
> Thanks for this review! Please see inline.
>
> > On Apr 12, 2016, at 4:06 PM, [e-mail] wrote:
> >
> > First, thanks very much to the authors Manav, Carlos, Sam, and Trilok
> for their work on this document.
> >
> > As is customary, I have done my AD review before requesting IETF Last
> Call.  In this case, I have a couple minor comments that I would like the
> authors to address during IETF Last Call.
> >
> > In addition to IETF Last Call, I am requesting a Routing Directorate
> review.  I expect that both of these will conclude by April 27 and that
> this draft will be on the IESG telechat on May 5.  During this period, it
> is critical that the authors be extremely responsive and update the draft
> as appropriate so that the process runs as smoothly and quickly as feasible
> >
> >
> > Minor comments:
> >    1) Draft references RFC 4970 instead of RFC 7770 which obsoleted it.
> In addition to updating the reference, please reread RFC 7770 and be
> certain that there are no surprises that can come from multiple RI LSAs
> being allowed or other nuances.  I personally don't see any right now.
>
> Great point. Working copy updated. I also re-read RFC 7770 and I do not
> think there’s any additional considerations or implications.
>
> >
> >    2) In Sec 2.1, it specifies "Routers that do not recognize the S-BFD
> Discriminator TLV Type MUST ignore the TLV."  I don't think that this
> document can mandate what routers that don't implement it do.  I went back
> through RFC 7770 and don't see any description *sigh* for the expected
> router behavior if a sub-TLV isn't recognized.  This might be a very useful
> errata to add to RFC 7770 - unless someone else can find where the behavior
> is specified.  For this draft, please think about what "ignoring the TLV"
> means and what routers that do not know about this draft are likely to do -
> and then update this sentence.
>
> True, this doc cannot say “MUST ignore”. However, RFC 7770 S2.3 already
> says “Unrecognized types are ignored.”
>
> I will change
> "Routers that do not recognize the S-BFD Discriminator TLV Type MUST
> ignore the TLV.”
> to:
> "Routers that do not recognize the S-BFD Discriminator TLV Type will
> ignore the TLV [RFC 7770], and therefore will not learn S-BFD
> Discriminators via OSPF.”
>
> Changes made in our working copy, and can submit when signaled.
>
> Feedback most welcome.
>
> Thanks,
>
> — Carlos.
>
> >
> > Thanks!
> > Alia
>
>
>