RE: Tanzanian ISPs move closer to establishing an Internet exchange

[Brian Longwe <brian at pch.net>]

In Kenya we started the IX when _AL_ absolutely _ALL_ the web-hosting was
on servers in the US.

Only mail was being exchanged initially but within three hours of peering
between four ISPs, Africa Online's 64k link was fried and went into
congestion collapse.


That's something called pent-up demand. Basically TCP/IP slows itself down
to accomodate the slow VSAT links that exchange traffic via overseas IXs,
but as soon as there is a faster pipe, TCP/IP speeds up and the effects
can be quite alarming if the rate of increase is interfered with by
bandwidth limitations....


Longwe

On Tue, 13 May 2003, Badru Ntege wrote:

> If I might add something to this.  I think we get too pre-ocuppied with the
> actual IX and forget the other variables that make it a viable commercial
> option.  An IX in a country that does not do any local hosting in the scheme
> of things has limited value. We have to look at other aspects that create
> local content and local value.  In most discussions I see about the IX's the
> value of creating local content seem to take a low priority.
>
> Untill there is local value it will always be of value for African nations
> to connect to the us and europe. (we need them more than they need Us)  It's
> amazing that even most african nations host their government websites in the
> US and Europe, our own communications commision in uganda is hosted in the
> US. Any idea where www.nepad.org is hosted ?  So what is the imperative to
> get a connection to other ISP's ?.
>
> We need to actually identify other commercial values than the fact that my
> email has to go to the US then Europe before it gets to the guy next door.
> If he is using a dial-up and his email takes an extra 10 mins to get to the
> ISP which he will check once a day what is the value.
>
> Please do not get me wrong I'm a strong proponent for IXP's but I believe we
> need to look at and sell the actual business value like business to business
> connections, local webhosting etc to actually give more tangible value to
> the IX.
>
> More and more the people who actually make the IT decisions especially in
> africa are not the generation that is IT aware some of these folks do not
> even have a computer on the desktops, most of them if they do it's cosmetic.
> Let's not forget that the Internet is not that old on the continent, so we
> need to actually deliver a message that can be understood.
>
> My worry is to get an IX in place and after some time no commercial value is
> realised, I know it's an uphill struggle but let us design the message for
> the audience and then build in all the other variables.  That way we will
> deliver a solution that works and is sustainable.
>
> badru
>
>
> -----Original Message-----
> From: owner-afnog at afnog.org [mailto:owner-afnog at afnog.org]On Behalf Of
> John Tully
> Sent: Tuesday, May 13, 2003 4:17 PM
> To: afnog at afnog.org
> Subject: Re: Tanzanian ISPs move closer to establishing an Internet
> exchange
>
>
>
> It seems a difficult number to estimate, but you can be sure that the
> dollar cost of not having a local IX in Africa is much much more expensive
> than the same issue for a US or European ISP.
>
> If you take into account the prices that the local telephone companies
> charge for leased lines to Europe and the European Internet prices, then
> the cost for bandwidth in Africa is many multiples of what US ISPs pay for
> bandwidth.  Plus, consider if you don't have a local IX, then the waist of
> bandwidth is going two ways -- all the way to Europe or the US and then
> back to the local server...
>
> John
>
>
>
> >On Tue, May 13, 2003 at 12:53:32PM +0000, ALAIN PATRICK AINA wrote:
> > >
> > >
> > > > could you give us the basis for the $400m figure?
> > >
> > > Do you also want to know the basis when WHO said " MALARIA KILLS ONE
> > PERSON
> > > EVERY 20 SECOND IN AFRICA"? ;-)
> >
> >That statistic can easily be checked: it equates to about 1.5m people dying
> >per year, which sounds plausible to me, and I trust the WHO to count the
> >death certificates.
> >
> >But the $400m figure seems ludicrous. Surely all the ISPs in Africa
> together
> >don't even spend that much on transit and link costs? And then only a small
> >fraction of that can be saved by peering.
> >
> >Regards,
> >
> >Brian.
> >
> >-----
> >This is the afnog mailing list, managed by Majordomo 1.94.5
> >
> >To send a message to this list, e-mail afnog at afnog.org
> >To send a request to majordomo, e-mail majordomo at afnog.org and put
> >your request in the body of the message (i.e use "help" for help)
> >
> >This list is maintained by owner-afnog at afnog.org
>
>
> -----
> This is the afnog mailing list, managed by Majordomo 1.94.5
>
> To send a message to this list, e-mail afnog at afnog.org
> To send a request to majordomo, e-mail majordomo at afnog.org and put
> your request in the body of the message (i.e use "help" for help)
>
> This list is maintained by owner-afnog at afnog.org
>
>
> -----
> This is the afnog mailing list, managed by Majordomo 1.94.5
>
> To send a message to this list, e-mail afnog at afnog.org
> To send a request to majordomo, e-mail majordomo at afnog.org and put
> your request in the body of the message (i.e use "help" for help)
>
> This list is maintained by owner-afnog at afnog.org
>


-----
This is the afnog mailing list, managed by Majordomo 1.94.5

To send a message to this list, e-mail afnog at afnog.org
To send a request to majordomo, e-mail majordomo at afnog.org and put
your request in the body of the message (i.e use "help" for help)

This list is maintained by owner-afnog at afnog.org