Search RPD Archives
Limit search to: Subject & Body Subject Author
Sort by:

[AFRINIC-rpd] New Policy Proposal: Inter RIR IPv4 Address Transfers (AFPUB-2013-V4-001-DRAFT-01)

Sunday Folayan sfolayan at gmail.com
Mon Jan 14 22:02:19 UTC 2013


On 14/01/2013 22:31, Andrew Alston wrote:
> Hi David,
>
>>> oh ... we can even estimate 3 addresses per students population, since
> they will have laptops, pads and fones, all requiring Wifi at the minimum.
>
>> I think it would be a mistake to propose a particular ratio in policy
> since that number will likely vary over time and circumstances (e.g., these
> days, a software/web developer will likely have a number of virtual machines
>
>> in addition to their laptop, tablet, and phone(s)). One size doesn't fit
> all. I'd hope it would be sufficient for requesters to provide the ratio and
> the justification for that ratio with their request.
>
> Just one comment on this, I to believe that specifying a ratio in an ideal
> world should not be necessary, however, history in applications from
> Universities has proven this is far from an ideal situation.  I refer to an
> application within the last 6 months, where the argument about ratios to of
> students to addresses took center stage for more than 2 weeks, with
> arguments coming back for example that "When a student is in a lab, working
> on a lab pc, they won't be using their phone, therefore it doesn't need an
> IP, therefore the ratio is reduced."

+1

>
> The 3 IP per student ratio however is something we've found to hold as a
> pretty good average, we found this by analyzing the wireless leases and
> other leases at two Universities in South Africa during this aforementioned
> discussion (ratio was actually 2.something, I forget exactly what), but 3
> allows for some growth.  While I agree that some people may have more
> devices/virtual machines than the three in the ratio, if you average out
> over an entire student population within an institution, I think it's a
> pretty good number.

Reduce complexity and subjectiveness is our objective. We are dealing 
with averages, not exceptions, so ... Agreed.

> One of the other points of the proposed policy which I'll be submitting as
> soon as I get Sunday's comments back on the exact wording, is to reduce the
> complexity around Academic institutions getting space.

Sent to you.

> As stated in
> previous emails, one of the major drivers of NAT in these universities and
> the resistance to getting more space is the bizarrely complex and time
> consuming process to get AfriNIC to actually allocate anything.  By using a
> student head count and a fixed ratio, we drastically reduce this complexity
> and the amount of paperwork needed.
>
> Policy proposal to follow shortly.

I guess we should do one more authors' iteration before we roll :)


Sunday.




More information about the RPD mailing list