> On Feb 17, 2025, at 09:05, Ketan Talaulikar <[email protected]> wrote:
> 
> < as a LSR WG member >
> 
> Hi Acee/All,
> 
> Please check inline below.
> 
> On Mon, Feb 17, 2025 at 7:16 PM Acee Lindem <[email protected] 
> <mailto:[email protected]>> wrote:
>> +Gunter and Ketan
>> 
>>> On Feb 17, 2025, at 07:27, Acee Lindem <[email protected] 
>>> <mailto:[email protected]>> wrote:
>>> 
>>> Hi John, 
>>> 
>>>> On Feb 16, 2025, at 8:16 PM, John Scudder via Datatracker 
>>>> <[email protected] <mailto:[email protected]>> wrote:
>>>> 
>>>> John Scudder has entered the following ballot position for
>>>> draft-ietf-lsr-ospf-admin-tags-26: No Objection
>>>> 
>>>> When responding, please keep the subject line intact and reply to all
>>>> email addresses included in the To and CC lines. (Feel free to cut this
>>>> introductory paragraph, however.)
>>>> 
>>>> 
>>>> Please refer to 
>>>> https://www.ietf.org/about/groups/iesg/statements/handling-ballot-positions/
>>>>  
>>>> for more information about how to handle DISCUSS and COMMENT positions.
>>>> 
>>>> 
>>>> The document, along with other ballot positions, can be found here:
>>>> https://datatracker.ietf.org/doc/draft-ietf-lsr-ospf-admin-tags/
>>>> 
>>>> 
>>>> 
>>>> ----------------------------------------------------------------------
>>>> COMMENT:
>>>> ----------------------------------------------------------------------
>>>> 
>>>> Thanks for this document. I have just a few questions and comments, below.
>>>> 
>>>> - In Section 4, "An OSPF router supporting this specification MAY be able 
>>>> to
>>>> advertise and propagate multiple tags." Does "propagate" have some 
>>>> well-known
>>>> meaning in OSPF that is different from "flood"? Because I assume that any 
>>>> OSPF
>>>> router MUST flood (propagate!) whatever it's given. (I do see paragraph 4,
>>>> which hints that the answer is "yes in OSPF 'propagate' always means 
>>>> 'between
>>>> areas'" but I'd appreciate a confirmation.)
>>> 
>>> Yes. This is the meaning of "propagate" in this context. 
>> 
>> I can clarify this on the first usage of “propagate”. 
>> 
>> -  specification MAY be able to advertise and propagate multiple tags. The
>> +  specification MAY be able to advertise prefixes with multiple tags and 
>> propagate prefixes
>> +  with multiple tags between areas. The
>> 
>> 
>> 
>>> 
>>> 
>>> 
>>>> 
>>>> - In Section 4, "Depending on the application, the number of tags 
>>>> supported by
>>>> the OSPF routers in the OSPF routing domain MAY limit deployment of that
>>>> application." That seems like a misuse of RFC 2119 MAY -- aren't you just
>>>> stating a possibility, not giving permission for an implementation choice?
>>> 
>>> I don't feel strongly. The statement is simply to indicate that an 
>>> implementation can support multiple tags but limit the number supported. 
>> 
>> I’d like to hear get Gunter’s and Ketan’s opinion as whether to change this 
>> from “MAY” to “may”. 
> 
> KT> The use of "may" seems appropriate to me

Yes - I mistakenly looked at another “MAY” in the document. For this limiting 
application usage, I agree and will change to “may”. 


>  
>> 
>> 
>>> 
>>> 
>>>> 
>>>> - In Section 4, "When tags are advertised for AS External or NSSA LSA 
>>>> prefixes,
>>>> the existing tag in the OSPFv2 and OSPFv3 AS-External-LSA and NSSA-LSA
>>>> encodings MUST be utilized for the first tag." I almost never do this, 
>>>> but...
>>>> couldn't this be a SHOULD? The case I can think of where you might want to 
>>>> use
>>>> the Administrative Tag Sub-TLV even as the first tag, is in some future 
>>>> time
>>>> when Administrative Tag Sub-TLV is ubiquitous and you are looking toward
>>>> deprecating the old encoding. I suppose a new RFC could be written at that
>>>> time, to permit the variance, but why not just allow it here with a SHOULD?
>>> 
>>> The thinking here was that we'd always support the existing support for a 
>>> single AS-External tag in OSPFv2 and OSPFv3 encodings
>>> The motivation just isn't that strong since in order to get to ubiquitous 
>>> deployment, implementations would need to support the
>>> existing tag advertisement. 
>> 
>> Also, like to get Gunter’s and Ketan’s opinion here. Again, I don’t think it 
>> would be worth the pain to ever require implementations to support the new 
>> encoding for the existing tag. 
> 
> KT> This text is referring to the usage of the base LSAs for both OSPFv2 and 
> v3 - as long as they are used (no option for v2, but there is a possibility 
> for v3 - see further) then for me the MUST in the current text is the right 
> approach for backward compatibility. Now, in OSPFv2, there is no proposal to 
> get rid of the based fixed LSAs - so it doesn't make sense to deprecate the 
> tag field in them. In OSPFv3, we have the base and then the extended LSAs 
> (RFC8362) - there is a path to switch OSPFv3 completely to the use of the 
> extended LSAs and stop using the base LSAs. In that E-LSA only case, for 
> OSPFv3 we will have the Route Tag sub-TLV for a single tag and this new 
> sub-TLV for multiple tags. So, for OSPFv3 we will have 3 ways. One can argue 
> for the deprecation of the Route Tag sub-TLV - but I am aware of 
> implementations. This is anyway orthogonal to the text blob that we are 
> discussing.

In any case, I think that if we were to deprecate usage solely for the OSPFv3 
Extended LSA case, it would require a separate document. I’m going to leave 
this as a “MUST”. 

Thanks,
Acee


> 
> Thanks,
> Ketan 
>  
>> 
>> Thanks,
>> Acee
>> 
>> 
>> 
>>> 
>>> Thanks, 
>>> Acee
>> 

_______________________________________________
Lsr mailing list -- [email protected]
To unsubscribe send an email to [email protected]

Reply via email to