Re: [OSPF] RtgDir review: draft-ietf-ospf-ospfv3-autoconfig-09

["Acee Lindem (acee)" <[e-mail]>]

Hi Martin, 

On 1/5/15, 7:07 AM, "Martin Vigoureux"
<[e-mail]> wrote:

>Hello,
>
>I have been selected as the Routing Directorate reviewer for this draft.
>The Routing Directorate seeks to review all routing or routing-related
>drafts as they pass through IETF last call and IESG review, and
>sometimes on special request. The purpose of the review is to provide
>assistance to the Routing ADs. For more information about the Routing
>Directorate, please see ​
>http://trac.tools.ietf.org/area/rtg/trac/wiki/RtgDir
>
>Although these comments are primarily for the use of the Routing ADs, it
>would be helpful if you could consider them along with any other IETF
>Last Call comments that you receive, and strive to resolve them through
>discussion or by updating the draft.
>
>Document: draft-ietf-ospf-ospfv3-autoconfig-09
>Reviewer: Martin Vigoureux
>Review Date: 2015-01-05
>IETF LC End Date: n/a
>Intended Status: Proposed Standard
>
>Summary:
>This Document is ready for publication. It has one or two typos.
>I have a couple of questions (see Minor Issues).
>
>Comments:
>This Document is well written and provides the necessary information and
>context for readers to understand what it specifies.
>
>Minor issues:
>    As OSPFv3 Router implementing this specification must select a unique
>
>Is that a must or a MUST? I guess it is a must since it is said
>afterwards that the uniqueness is not 100% guaranteed, but I just wanted
>to make sure.
>Yet, since there is a possibility of a Router ID collision, couldn't the
>sentence be rephrased as follows to reflect the reality:
>    An OSPFv3 Router implementing this specification must ideally select
>    a unique Router ID.

Ultimately, an OSPFv3 router MUST have a unique OSPFv3 Router-ID (although
this may require multiple attempts at selection). How about this?

    An OSPFv3 router requires a unique Router-ID for correct protocol
operation. An OSPFv3 router implementing this specification will select an
initial Router-ID with a high probability of uniqueness. ….





>
>
>    An OSPFv3 router implementing this specification MUST compare a
>    received self-originated Auto-Configuration LSA's Router-Hardware
>    Fingerprint TLV against its own router hardware fingerprint. If the
>    fingerprints are not equal, there is a duplicate Router ID conflict
>    and the OSPFv3 Router with the numerically smaller router hardware
>    fingerprint MUST select a new Router ID as described in Section 7.3.
>
>I feel that these two sentences are not crystal clear. Forgive me if it
>is only due to me not being a native English reader.
>The second sentence implies that fingerprints between "a received
>self-originated" LSA and a router's own hw fingerprint can be different.
>In the first sentence, I read "self" as referring to the router which is
>the subject of that sentence, and I therefore fail to understand how an
>LSA originated by a router could arrive back to that router but with a
>different fingerprint.
>Also, the second sentence seems to imply that iff fingerprints are
>different then the Router IDs are the same. I know that we are in a
>section about Duplicate Router ID, but just as for Section 7.1 which
>clearly sets the conditions, it might be worth saying that if
>fingerprints are different and OSPFv3 Router IDs identical then there is
>a duplicate Router ID conflict. But again, this might not be needed if
>my reading of *self* is wrong.

I agree that this is not clear since it relies on one deducing that an LSA
is perceived as self-originated if it has the same Router ID. How about
this: 

An OSPFv3 router implementing this specification MUST detect received
Auto-Configuration LSAs with its Router ID specified in the LSA header.
LSAs received with the local OSPFv3 Router's Router ID in the LSA header
are perceived as self-originated (see section 4.6 of [OSPFV3]).  In these
received Auto-Configuration LSAs, the Router-Hardware-Fingerprint TLV is
compared against the OSPFv3 Router's own router hardware fingerprint.  If
the fingerprints are not equal, there is a duplicate Router ID conflict
and the OSPFv3 router with the numerically smaller router hardware
fingerprint MUST select a new Router ID as described in Section 7.3.





>
>
>Nits:
>    OSPFv3 SHOULD be auto-configured on for IPv6 on all interfaces
>Isn't "on" after "auto-configured" superfluous?

Fixed. 

>
>s/As OSPFv3 Router implementing/An OSPFv3 Router implementing/

Fixed. 

>
>The document uses both "OSPFv3 Router" and "OSPFv3 router".
>It may be worth using only one way of writing it.

These should be "OSPFv3 router”

Thanks,
Acee 

>
>
>Thanks
>
>Martin,
>wishing you all the best for 2015