<html>
<head>
<meta content="text/html; charset=UTF-8" http-equiv="Content-Type">
</head>
<body bgcolor="#FFFFFF" text="#000000">
<div class="moz-cite-prefix">[Forgive the length of this email]<br>
<br>
On 28/01/2013 15:54, Nii Narku Quaynor wrote:<br>
</div>
<blockquote
cite="mid:83FE465A-3C31-426B-B0F7-05D698D8BBA6@ghana.com"
type="cite">
<meta http-equiv="Context-Type" content="text/html; charset=utf-8">
<div>And the university does not have a network diagram? and
implementing a network <br>
</div>
</blockquote>
<br>
There definitely will be a Network plan! The real sketches end at
the core. The remaining parts of the plan are assumptions, not
things cast in stone.<br>
<br>
How many IP's will be needed for concurrent connections? ... our
initial headache! Andrew could not convince the hostmasters that he
needed 1.8 IPs per student.<br>
<br>
University A's plan may show ... 50 concurrent connections per
wireless AP<br>
University B's plan may show ... 200 concurrent connections per
wireless AP<br>
<br>
These connections could be any or all of photabtops. What other
justifications can be included in diagrams, I mean, what other
diagram/shape (beyond the figures) will differentiate 50 wireless
users from 200 wireless users within the last mile?<br>
<br>
This is the real challenge we should not ignore in all the muses
Andrew was pushing, I would say sometimes with energy and a rasp
file, before I offered to co-author the policy with him.<br>
<br>
<blockquote
cite="mid:83FE465A-3C31-426B-B0F7-05D698D8BBA6@ghana.com"
type="cite">
<div>The information, referenced as at start, was about design
options not about a 'network plan' being reviewed</div>
<div><br>
</div>
<div>These kinds of options and issues also apply to any network
and may not be peculiar to universities. In the end you have a
design and resource plan<br>
<br>
</div>
</blockquote>
<blockquote
cite="mid:83FE465A-3C31-426B-B0F7-05D698D8BBA6@ghana.com"
type="cite">
<div>Ok, fix it but a fix where <span class="Apple-style-span">is
needed. I wonder if a fix by policy is what's needed in this
case</span></div>
<div><span class="Apple-style-span"><br>
</span></div>
</blockquote>
<br>
In moving from design options to network plans, Our goal is to (1)
keep it simple for Academic institutions. (2) legitimately Increase
the burn rate of IPs in our region. If there is a simpler way of
doing this without subjecting it to the ingenuity of the applicants,
or the intuition of the hostmaster (at least for Academic
institutions as captioned in the proposal) I am willing to withdraw
the policy (or at least my Name as Co-Author).<br>
<br>
<blockquote
cite="mid:83FE465A-3C31-426B-B0F7-05D698D8BBA6@ghana.com"
type="cite">
<div><span class="Apple-style-span">We are actually imposing
constraints eg end-user category, now 5:1 expectation, etc</span></div>
</blockquote>
<br>
I don't think so Nii. I would say ... not more than 5:1 expectation
without justification. Anything above that, must be justified. And I
must point out that this policy will die a natural death in less
than 2 years! <br>
<br>
AfriNIC is not the only stakeholder that wants some form of
justification/convincing before allocating resources. The ability of
the Universities ICT personnel to convince their management that
indeed there are up to 3,4 or 5 devices per students to make them
pay more for a higher category is the limiting real factor here.<br>
<br>
<br>
Sunday.<br>
<br>
</body>
</html>