wiki:QA review Sep 2014

Routing Directorate Quality Assurance (QA) review

draft-filsfils-spring-segment-routing-mpls-03

Summary

I have some concerns about this document that I think could be resolved after the adoption as a working document. I also have one question (see below) that I would like to have answered before adoption as a working group docunment.

Comments

From a tecchnical point of view I find the document (given a response on my question on domain wide labels) ready to be adopted as a wroking group document.

I have in other reviews said that I don't theink I'm competent to do an accurate language review. For this document however I find it a bit (lack of better word) declarative/proclamative, i.e. it use a language that sometimes is a bit too close to marketing language. I think this could be addressed after the document has been adapted as a working group draft.

I think it would have been good if this adoption poll had been copied to the mpls working mailing list, at least the wglc need to be.

Major Issues

No major issues found.

Question

There has been some confusion if this document (and the other SPRING document) actually proposes domain wide labels. Domain wide labels has been proposed several times before and has been rejected and is thought to contrary to the MPLS architecture.

When applied to MPLS, a Segment is a LSP and the 20 right-most bits of the segment are encoded as a label. This implies that, in the MPLS instantiation, the SID values are allocated within a reduced 20-bit space out of the 32-bit SID space.

The next section says:

The notion of indexed global segment fits the MPLS architecture [RFC3031] as the absolute value allocated to any segment (global or local) can be managed by a local allocation process (similarly to other MPLS signaling protocols).

If the segement is a domain, and the domain is encoded as a label, then it is possible to read this as that SID is a domain wide label.

I don't think that is the case but the document is somewhat ambigious.

I think a clear statement that this draft is not entertain domian wide labels is needed.

On the other hand (if I'm wrong) and this really is a domain wide label, the decsion whether domain wide labels is OK should be taken in cooperation with the MPLS wg.

Minor Issues


Number of authors on the fist page

The RFC editor guidelines are clear:

"A small set of author names, with affiliations, may appear on the front page header. These should be the

lead author(s) who are most responsible for the actual text. When there are many contributors, the best

choice will be to list the person or (few) persons who acted as document editor(s) (e.g.,"Tom Smith,

Editor"). There is no rigid limit on the size of this set, but there is likely to be a discussion if the set exceeds

five authors, in which case the right answer is probably to list one editor." 12 authors is too far from the recommended 5. See: https://www.rfc-editor.org/policy.html#policy.authlist This need to be fixed before wglc, but should be fixed as soon as possible.


Introduction

While the introduction includes very useful information on where to find the rest of the SR realted documents, there is very little infor on what this document actually contribute.

I think the introduction should at least address what the problem that the document addresses is.


Type of document

The document says Standards Track. Exactly why does this need to be Standards Track?


Nits

Abbreviations

A list of abbreviation would be useful.


Well-known abbreviations

I've not checked on all abbreviations, but SID is (not yet a wellknow abbreviation) and need to be expanded. Or rather "SID" is but it is expanded in more than one way and none stands for Segement Identifier. Please check that all abbreviations are expanded at first use or is one of tje RFC Editors wellknow abbreviations. Note to wg chairs: This seems to a problem shared with several SR drafts.


IGP's

The document use IGP to be synomous with IS-IS and OSPF, but strictly there arr more IGP's, make the scoping to IS-IS and OSPF explicit.


Last modified 5 years ago Last modified on Sep 29, 2014, 3:44:54 AM