[AFRINIC-rpd] Academic IPv4 Allocation Policy Second Draft (AFPUB-2013-GEN-001-DRAFT-02)

[Andrew Alston]

Hi SM,

We have discussed within AfriNIC at various times over the years holding
onto our space so that it will be available as we develop.  The same applies
to Universities.  I would argue that any University with 10 thousand
students to remain competitive will upgrade their infrastructure, and that's
happening at a drastically accelerated rate, because the internet
infrastructure on university campuses is becoming critical to the academic
experience (as stated by Guy Halse).

Furthermore, since it is fast becoming cheaper for a University to upgrade
its infrastructure to support such access than to deliver content via more
traditional methods, the progression of infrastructure is a natural thing.
I point at an example of a small University in South Africa that is
predominantly rural, and actually generally runs very close to the red in
financial terms.  Because of a need to drive down costs and provide better
access, they, even in their financial position, realized it was prudent to
do massive upgrades, and went and sourced the money to do so.  The
Universities may not upgrade at the same rate, but within the next 3 years,
because of the criticality of online resources to teaching and academia,
wifi network infrastructure to allow students to access things *will* be
common place in my opinion.

So, in answer to question (a), no, the non-utilization problem will resolve
itself shortly, and furthermore, there is no actual practical way to audit
this anyway and (b) no, that is something that will happen over the course
of time.

Look, I had a long discussion last night telephonically with Nii about this,
and I would love to see a situation where a University could specify a ratio
they *needed* up to a maximum of X, so if we said a university can turn
around and say we need a ratio of 3:1, whereas another can say we need 5:1,
both are <= 5, both are accepted by AfriNIC on face value, I would accept
this.  However, AfriNIC would then need to commit to accepting these values
at face value argument and back and forth.  Otherwise, we need the fixed
ratio in policy to avoid the same issues we are currently having.  

Currently the policy sets a minimum of 5, higher can be justified.  I would
be happy to say any ratio below or equal to 5 is accepted de-facto based on
institutional statement, anything above 5 requires further justification
that can be investigated.  Would this suffice?

Andrew


-----Original Message-----
From: SM [mailto:sm at resistor.net] 
Sent: Saturday, January 26, 2013 11:07 AM
To: Badru Ntege; Andrew Alston
Cc: rpd at afrinic.net
Subject: Re: [AFRINIC-rpd] Academic IPv4 Allocation Policy Second Draft
(AFPUB-2013-GEN-001-DRAFT-02)

Hi Badru,
At 20:18 25-01-2013, Badru Ntege wrote:
>The statistics are showing the demand for the urban institutions with 
>the financial resources to build a network that can support this demand 
>and thus the need to have an idea of the network.  But out in the real 
>world not all academic institutions have the resources to build such 
>networks.
>
>which is why it is a good idea to have an understanding of the network 
>size.  if we go by institutional population we might be introducing a 
>major flaw.

I'll try and restate what you said in the first 
paragraph.  University A and B each have 10,000 persons.   University 
A has the resources to build a network to service these 10,000 persons.
University B does not have the resources to build a network to service these
10,000 persons.  If I use population count as a measure I would be giving
50,000 IP addresses to University A and
50,000 IP addresses University B.

University B is using 5,000 IP addresses only and 45,000 IP addresses are
not utilized.  The amount of free IP addresses for the region reaches a
level where I cannot provide IP addresses to University C which has a
population of 5,000 persons.  Somebody points out that there are 45,000 IP
addresses not being utilized at University B.  I cannot do anything about
it.

The questions are:

  (a) Should I fix the non-utilization problem?

  (b) Is it possible for me to fix the non-utilization problem?


Regards,
-sm