Re: EPP host renames

On 20/12/12 07:08, Patrik Fältström wrote:
>
> On 19 dec 2012, at 19:47, Seth Goldman <[email protected]> wrote:
>
>> Isn't a hole in the EPP spec then?
>
> There are a number of unclear things in the EPP spec when you combine domain
> and host objects plus think of who the sponsoring registrar is.
>
> I have a case where this is the scenario in one TLD:
>
> Registrar A is sponsor for examplea.tld. Registrar B is sponsor for
> exampleb.tld. Now examplea.tld want to use ns2.exampleb.tld as name server. A
> name server that exists in the zone for exampleb.tld but not in the tld zone
> as glue (yet). The problem is that Registrar A is not allowed to create the
> ns2.exampleb.tld host object as they are not sponsor of the exampleb.tle, and
> it is not allowed to update the domain object exampleb.tld as the host object
> is subordinate to tld. A deadlock that so far after several months is not
> possible to resolve even via out of band mechanisms (i.e. pen, paper and
> letters).
>
> So, having host objects as real objects can create very very very problematic
> situations. So be careful what you ask for.
>
> Patrik
>

Hi,

I completely agree, but how realistic is the scenario? At the end, the zone of 
examplea.tld must be published on the actual name server addressed via 
ns2.exampleb.tld, i.e. the name server must be configured as a master or slave 
respectively, so it is likely that the two parties have to communicate somehow 
anyway. Only if more than two parties are involved, then probability of such a 
scenario increases IMHO.

Regards,

Klaus

_______________________________________________
provreg mailing list
[email protected]
https://www.ietf.org/mailman/listinfo/provreg