[RZERC] Possible scenarios to take up or turn down
Howard Eland
heland at afilias.info
Tue Aug 22 15:27:19 UTC 2017
> On Aug 22, 2017, at 9:44 AM, Howard Eland <heland at afilias.info> wrote:
>
>
>> On Aug 21, 2017, at 4:39 PM, Kim Davies <kim.davies at iana.org> wrote:
>>
>> Hi folks,
>>
>> Quoting Jim Reid and others on Monday August 21, 2017:
>>>>
>>>> - Introduction of RRTYPEs not previously included in the root zone (this would cover DNSSEC in the past).
>>>> - Changes which would cause a “significant” change in size to the root zone (Significant TBD).
>>>> - Significant changes to the operation of the root zone, such as
>>>> - Change in root zone operator
>>
>> What do we envisage RZERC's role is in this scenario?
>
> Perhaps a review of any new potential operator’s history as it pertains to DNS operations? Impact study of how the new operator’s network coverage could potentially change query paths?
>
>>
>>
>>>> - Change in DNSSEC key algorithm or size
>>>> - KSK roll?
>>
>> Assuming this KSK roll concludes in a normal fashion, it seems to me
>> performing a key rollover will become perfunctory (i.e. rolling every x
>> years). I don't see a KSK roll being elevated to this group unless
>> it meets the other criteria above (i.e. new algorithm, possibly new
>> size.)
>
> What about emergency KSK rolls? I’m not saying we bless a roll before it occurs, but should we look at the procedures for this type of activity to ensure they are in line with keeping the stability of the root zone?
>
>>
>>>> - Changes in operational policy
>>>> - Significant change in root DPS (Significant TBD)
>>>>
>>>>
>>>> Possible list of topics the RZERC committee should decline to review:
>>>>
>>>> - Daily “standard” changes to the root zone (Standard TBD).
>>>> - Introduction of new TLDs into the root zone (subject to the size bullet above).
>>
>> A couple of other scenarios I had in mind that don't reflect architectural
>> changes to the root zone content itself:
>>
>> - Support for additional algorithms and digest types for delegations.
>
> I believe I mentioned this above for the root zone, but yes, the same could be said for delegations.
>
>> - Upgrades to the software used to manage the root zone and root zone workflow
>> (except as it pertains to other characteristics deemed to be in-scope)
>
> Not sure I want to do work every time there’s a patch to an operator’s software. Would that also include network gear upgrades? This seems intractable to me.
Please ignore my nonsense immediately above. I mis-read Kim’s list as the “yes” list, not the “no” list. Apologies to all of the confusion (and thanks to Peter for quietly pointing out my error!). I agree this is out of scope.
>
>> - Revisions to the technical check criteria for the root zone.
>
> +1
As part of my misread - I actually think we should be looking at this.
>
>> - Revisions to the authentication mechanisms for TLD managers.
>
> Not sure exactly what you mean on this last one.
Since I’m not sure what this is, I can’t say one way or the other here.
>
>>
>>> One thing that I think could/should be added to the list is getting
>>> a root key repository outside the USA. Suppose US airspace gets closed
>>> down. Or immigration policies make it impractical for foreigners to
>>> serve as TCRs.
>>
>> What would be the scope of RZERC's consideration of such a change? Where
>> the root management partners keep their physical facilities seems to me
>> beyond "architectural changes to the content of the DNS root zone”.
>
> I’m concerned that this moves us more into Layer 9 - are we going to look into International policy changes?
>
>>
>>> I assume that RZERC is not going to do any work on whatever ends up
>>> on the list we compile. Until of course someone approaches us with a
>>> proposal that’s in scope for RZERC.
>>
>> I think that's correct. What we discussed on the call was having illustrative
>> scenarios that would inform thinking about what is in-scope and out-of-scope.
>
> +1
>
> -Howard
>
>>
>> kim
>
More information about the RZERC
mailing list