[afnog] Challenges in the African Continent

[SM]

Hi Graham,
At 02:09 31-03-2009, Graham Beneke wrote:
>No - I am referring to smaller scale deployments as well. I have come
>across a number of small/micro ISPs on the continent that are running
>hosting operations that are not up to standard:
>
>* Using underpowered hardware that is unable to serve requests timeously.
>
>* No useful power backup. A consumer grade 600VA UPS that lasts 20
>minutes does not count as power backup.
>
>* No network redundancy either with local hardware or upstream
>providers. And often - over contended upstreams.

The above description (except for the 600VA UPS) can apply to hosting 
in several "world class" datacenters.  You'll only find out when 
things go wrong.

>This in combination with largely inadequate peering infrastructures in
>many parts of Africa make 'local' hosting pretty pointless. With your
>traffic 'local' to your own customers and everyone else has to access it
>over 2 birds making the latency _double_ that of the internationally
>hosted equivalent.

I won't call "local" hosting pointless.  Some people choose that 
because they understand how it will benefit them in the long run.  It 
can also be easier to reach someone local than a provider in a far 
away country.  The effects on the industry are:

  1. it increases local know-how.  That can turn into job creation.

  2. there is vertical integration

  3. you are closer to the end-user (keep local traffic local)

If you want to speed up international access, you can always redirect 
that traffic to an external site.  That's quite easy to implement.

Mark mentioned that there is underutilized infrastructure that could 
be used for hosting.   These sites must have to some level of 
reliability or else we would not have had 24/7 access to the Internet 
in the region.

At 01:59 31-03-2009, Graham Beneke wrote:
>Why is it that the Americans, Europeans and Chinese can see the business
>benefits of investing in Africa and are spending billions of their own
>money doing so. Meanwhile - many of us in Africa are still looking for
>the business case...

It's the same reason why your neighbour earns more money than 
you.  He/she may see the opportunity and figure out how to derive 
revenue from it.  Building an infrastructure requires a huge 
investment.  Africans can be at a disadvantage when it comes to 
getting the capital.  It's easier to get money when you have 
money.  A multinational can leverage its existing infrastructure when 
it enters into new markets.  It can afford to make a loss while 
cross-subsidizing its other units.

I could say that there is a business case for investing in 
Africa.  Would that help Africa?  I don't think so.  I need market 
projections and real figures to back my claim.

Regards,
-sm