[lisp] Re: Rtgdir early review of draft-ietf-lisp-geo-06

[Luigi Iannone <[e-mail]>]

Hi Ines,

Do you consider that the document is now ready for publication from the RTGDir  perspective?

Thanks

Ciao

L.

> On 4 Jul 2024, at 07:35, Ines Robles <[e-mail]> wrote:
> 
> Ok, Thank you Dino for the clarification,
> 
> BR, 
> Ines
> 
> On Thu, Jul 4, 2024 at 1:39 AM Dino Farinacci <[e-mail] <mailto:[e-mail]>> wrote:
>> Ines, thanks for your comments. Here is one response to your commentary.
>> 
>> > > 9- In the security considerations, what about add description on attacks
>> > > related to geo-coordinates such as location spoofing?
>> > 
>> > We had added that from previous reviews. Tell us exactly what you are looking for.
>> > 
>> > Ok, thanks. I was wondering about potential consequences of location spoofing within the LISP environment, such as misleading network path selection. What do you think? 
>> 
>> I think we have covered this and there is no way to validate a "good geo-coordinate". If you authenticate the source who registered the mapping, you are trusting them. There is no way to do a back-door check to see if the location is correct or precise. 
>> 
>> We don't want the draft to spec out to validate something this:
>> 
>> EID: London, UK
>> RLOC: geo lat: 37, geo long: -121
>> 
>> Meaning, you don't want to say, "hey those coordinates are in San Jose, CA but you used a name called London, this is suspect, we probably shouldn't register this".
>> 
>> This sort of validation should be done in the implementation at the source (and not the LISP implementation) but the admins who decide London needs to be San Jose. ;-)
>> 
>> Dino
>> 
>> 
>>