Re: [DNSOP] Meeting feedbacks/summary on draft-bretelle-dnsop-recursive-iprange-location

[Manu Bretelle <[e-mail]>]

Thanks both,

So, I seem to gather that the main problem is to put forward Geolocation as
a way to return pseudo-targeted answers to end-users by using the resolver
IP as a proxy for it. This was more meant to be a use-case as to how geo
location has been used, but It is by no means my intent to push this as a
recommended solution. I am happy to drop this or rephrase.

I do believe that Geo can have other useful properties as Brian has
highlighted, even more so in an infrastructure that heavily relies on
anycast. I suppose in most cases one could find other approaches to get
that information whether the traffic coming from a given network is
legitimate, but in many cases, this will involve having access/knowledge to
the full stack (routing tables, peering links, AS paths...) which may not
be readily available to the ones operating DNS. Likely geo is not a perfect
proxy, but it does provide high enough signal that one can leverage.

Manu



On Fri, Nov 20, 2020 at 9:49 AM Brian Dickson <[e-mail]>
wrote:

>
> On Thu, Nov 19, 2020 at 9:35 AM Ben Schwartz <bemasc=
> [e-mail]> wrote:
>
>>
>> I do not support the geolocation function.  I think the right solution
>> here is ECS.  Even bad IP-geolocation from ECS will be better than using
>> the recursive resolver's country-code; at least it will be estimating the
>> location of the right entity.
>>
>> If ECS is not widely deployed enough for your purposes, I would focus on
>> ways to increase support for ECS.  For example, we could work on ways
>> to use shorter prefixes for improved privacy where appropriate, or provide
>> guidance on how to use ECS for geo-based responses.
>>
>
> I agree with the sentiment against providing differentiated answers based
> on the geo of the resolver (mostly).
>
> However, geo has other valuable uses.
> Thus, I'd like to retain geo as part of the standard.
> I think recommending against using it as a proxy for ECS should be
> included (with explanation, obviously).
>
> One reason to have explicit geolocation is that the geoip of the
> resolver's IP may not be accurate or timely, and for these use cases,
> explicit geo info is always better.
>
> One use case I raised at the meeting at the mic was grouping providers'
> regional resolvers for various purposes on the authority servers' side.
> Those things would include treating a regional group the same as a single
> resolver, for purposes of white-listing, or rate-limiting, or other
> operational activity.
>
> There are other kinds of things that authority operators would be able to
> do, some of which cannot be done today.
> Those include:
>
>    - Alerting on changes in affinity between resolver groups and
>    authority instances (which might indicate a routing problem or outage)
>    - Visibility on new deployments of resolver groups by a particular
>    resolver operator (which might benefit from resource allocation work on the
>    authority operator)
>    - Ability to detect or prevent spoofing of resolver IP addresses
>    - Ability to do cooperative work around ADoT
>    - Ability to correlate oddities (e.g. unexpected DNS cookies, set by
>    auth location A but seen by auth location B, associated with IP and
>    resolver Geo info)
>
> On the other hand, I could see a potential longer-term replacement of ECS
> with ECG (EDNS client geo) once the geo stuff is standardized.
>
> But let's not get distracted by that in the short term, where direct use
> of resolver geo has real value.
>
> Brian
>