For whatever it’s worth, during my own review I noticed the introduced fields 
and came to the same conclusion Dino does here, that it’s OK as written.

$0.02,

—John

> On Feb 12, 2023, at 8:46 PM, Dino Farinacci <[email protected]> wrote:
> 
> 
> 
> Okay I see the confusion and how this could be misleading. I’m not sure how 
> to fix this editorially.
> 
> The 2 fields are in a “Map-Reply Record” which was only in a Map-Register. If 
> a Map-Request would want to supply mapping entry, it would include a 
> Map-Reply Record. But before pubsub was spec’ed there would be no way to 
> encode the 2 new fields because the I-bit was not specified. 
> 
> Since the pubsub spec introduces the I-bit the 2 fields can be included and 
> needed for the new protocol operation sped ‘ed in the pubsub draft. 
> 
> A possible fix is to have pubsub refer to 9301, section 5.6 but would be 
> misleading to convey a Map-Register message which is not the intent. So I 
> conclude no change should be made. 
> 
> Dino
> 
>> On Feb 12, 2023, at 4:59 PM, Erik Kline <[email protected]> wrote:
>> 
>> 
>> On Sun, Feb 12, 2023 at 2:46 PM Dino Farinacci <[email protected]> wrote:
>> 
>>> The Map-Request registry can point to both 9301 and the new LISP PubSub RFC.
>>> 
>>> That works, yes.
>>> 
>>> I was wondering about the fact that the message itself just grew an extra 2 
>>> fields.
>> 
>> It shouldn’t have. 
>> 
>> Which fields are you referring to? If you are referring to site-ID and 
>> xTR-ID, those are existing fields in the Map-Register message (and not the 
>> Mal-Request message). 
>> 
>> I'm referring to the xTR-ID field and Site-ID field, both of which appear to 
>> be described as being "added to the Map-Request message defined in Section 
>> 5.2 of [RFC9301]", per Section 4 of the draft.
_______________________________________________
lisp mailing list
[email protected]
https://www.ietf.org/mailman/listinfo/lisp

Reply via email to