Re: [Last-Call] Opsdir telechat review of draft-ietf-pce-pcep-stateful-pce-gmpls-22

[John Scudder <[e-mail]>]

Thanks for your review, Sue.

Authors, thanks for all your work. Before we send the document to the RFC Editor, please consider whether to make changes based on the OPSDIR review. I agree with the points Sue has made; points 2 and 3 are about the description of MAY clauses and there is room for debate about them, so although I’d be happy to see clarification in these cases, I would accept the answer “we thought about this but decided not to make changes because _____”. Point 1, however, identifies a bug in the document which should be fixed (although note it is actually Section 7.1 that the comment applies to, not Section 7.2 as written). I’m not certain if the fix Sue proposes is the right one, or if the right fix is to change “if the PCE understands” to “if the PCC understands” at the beginning of the paragraph, but either way the paragraph as it stands doesn’t appear to make sense — I leave it to you to determine the correct fix.

Thanks,

—John

> On Jun 14, 2023, at 5:34 PM, Susan Hares via Datatracker <[e-mail]> wrote:
> 
> 
> Reviewer: Susan Hares
> Review result: Has Nits
> 
> OPS-DIR review:
> Status: Ready with nits
> General comment: Well-written, concise, and clear. The authors did excellent
> work.
> 
> NITS - in the error handling (sections 7.2, 7.3, and 7.3)
> 
> 1. section 7.2, paragraph 3, last sentence starting "Such a PCC"
> Old/Such a PCC MAY decide to utilize the capability even though it did not
> advertise support for it./
> 
> It seems as though PCC in this sentence is a typographical error for PCE.
> 
> 2. section 7.2, sentence 2,
> It is unclear what happens if the "MAY clause is taken".  Does the Stateful PCE
> simply report success or is it silent or does the error report get made?
> 
> Current text:/ If no LSP state information is available to carry out
> re-optimization, the stateful PCE SHOULD report the error "LSP state
> information unavailable for the LSP re-optimization" (Error Type = 19, Error
> value= TBD6), although such a PCE MAY consider the re-optimization to have
> successfully completed./
> 
> 3. section 7.3. sentence 5, beginning "An implementation May choose to"
> text: /An implementation MAY choose to ignore the requested exclusion when the
> LSP cannot be found because it could claim it that it has avoided using all
> resources associated with an LSP that doesn't exist./
> 
> Does the implementation send an error report in this case or does it positively
> ack the response? The MAY clauses lack clear actions.
> 
> Thanks for your hard work on this document.
> 
> 
>