[DT-F] REVISED: Design Team F kickoff

[Alan Greenberg]

At 12/04/2015 07:06 PM, Suzanne Woolf wrote:
>Alan,
>
>On Apr 12, 2015, at 2:31 PM, Alan Greenberg <alan.greenberg at mcgill.ca> wrote:
>
> > Sorry, I have had a number of personal functions this weekend 
> which have keep me away from e-mail.
> >
> > What we WILL accomplish by Thursday clearly remains to be seen. 
> What we were charged with accomplishing was to recommend changes to 
> the Root Zone management process as needed or desired for the 
> transition to take place.
> >
> > Currently the NTIA must approve all changes before they are 
> released to be RZ Maintainer (Verisign) for implementation.
> >
> > Since the NTIA will now be "out of the loop", we need to recommend:
> >
> > - How the authorization function is to be replaces, if at all
>
>Like Milton, I thought we'd been told it's not to be replaced.

Yes, you are correct. I guess I was being overlay cautious in case we 
discover something to cause that to be changed.


> > - What other changes do we recommend, if any, to compensate for 
> the removal of the NTIA authorization
>
>See my previous message-- I suggest we need to call out operational 
>impacts of NTIA's stepping aside more generally. Non-operational 
>aspects, i.e. the accountability enforced by having the contract in 
>the first place, are out of scope for us; but I tried to suggest a 
>couple of other factors we should look at, which may be outside the 
>day-to-day operation (the authorization function) but are still 
>relevant to IANA staff being able to do their jobs promptly and correctly.
>
> > - What other improvements do we recommend making to address 
> perceived weaknesses in the present process (despite the NTIA presence).
>
>I'm not sure I follow. I seem to recall that we've been told a 
>number of times now that people are happy with IANA's performance of 
>its tasks. Encoding the current service expectations outside of the 
>NTIA contract, so that they can continue to be met when the NTIA 
>contract is gone, is the work of the SLE group, isn't it?
>
>If that's incorrect-- could you clarify what "perceived weaknesses" 
>we need to address?
>
>If it's roughly correct-- it seems to me that our role is largely a 
>gap analysis: what dependencies on the NTIA contract for the root 
>zone management function are *not* addressed by other design teams 
>or the CCWG-Accountability? And what requirements are imposed on the 
>other design teams or the CCWG-Accountability?
>
>(I come to this task with a very operational orientation, so I 
>assume that there's a specific set of tasks that IANA and the RZM 
>partners need to perform, with or without the NTIA, in order to 
>carry out their responsibility for a correct, up-to-date production 
>root zone; and that we're making sure both that IANA and RZM staff 
>can adjust their procedures so they still know what they're supposed 
>to do in the absence of the NTIA contract, and that those 
>responsible for oversight post-NTIA also understand their 
>responsibility in detail, and can carry it out.)

Earlier messages had identified "weaknesses" related to checks and 
balances implemented largely by contractual terms. For instance, the 
RZ Maintainer *could* implement and publish changes different from 
those supplied by IANA. It has rarely happened (only one case that I 
have seen documented) but the potential is there. I am not taking 
thhe posiiton that we must address them in this group or even later, 
but they have been mentioned.


> > The NTIA has made it clear that any proposal officially submitted 
> must be ready for implementation - no further design 
> post-acceptance, so we MUST cover these areas before formal submission.
>
>I think I have the same perspective on this that Milton does-- other 
>operational communities tagged a couple of areas as subject to later 
>implementation, turned in their proposals, and are working on the 
>changes they identified as required between the initial proposal and 
>transition. The CWG doesn't have quite as much time, but it still 
>seems that we could do the same thing. (For example, we have to 
>identify who has to update software so that the RZM workflow is 
>correct without NTIA in the loop-- but there's no chance that will 
>be done by Thursday, or even by June, and no need to do it on that schedule.)
>
>If there's serious uncertainty about this, do we need to ask the ICG?
>
> > We have been requested to flesh this out as much as possible in 
> time for the publication of the next report, with an absolute 
> cutoff by this coming Thursday.
> >
> > I am hoping that we can make such a proposal, and you will see 
> more from me in this later today.
> >
> > In my mind, we do not have to reach closure on such a proposal, 
> but we should have SOMETHING for the communities to comment on.
>
>This seems like a reasonable goal if we're careful about what we're tackling.
>
>I'm planning to be on the phone tomorrow morning. I will not be 
>available tomorrow afternoon (US Eastern).
>
>
>thanks,
>Suzanne