[8731] in Commercialization & Privatization of the Internet
Re: NAP philosophy & IP technology
daemon@ATHENA.MIT.EDU (Erik E. Fair" (Your Friendly Postm)
Thu Dec 2 01:26:11 1993
From: "Erik E. Fair" (Your Friendly Postmaster) <fair@apple.com>
In-Reply-To: <9312020247.AA17063@psi.com>
To: "Roger Fajman" <RAF@cu.nih.gov>
Cc: com-priv@psi.com
Date: Wed, 01 Dec 93 22:25:26 -0800
Roger,
The process that I described for balancing Apple's Internet
connectivity is the very same process that any professional network
manager would go through to balance the load on her LAN/MAN/WAN. There
are just two differences with the Internet case:
1. The Internet is between two and three orders of magnitude
larger the typical business network, and
2. I don't control the whole thing. Not even close. I have to
negotiate and react, rather than being able to dictate.
These differences just make the job a little harder. Most of the
complexity comes from the wide array of choices one has for routing
protocols to calculate & move IP routing information around. Despite
appearances, this is a feature, not a bug. Most LAN protocol suites
have one routing protocol, period, and you'd better like it.
It is also worth mentioning, though, that the typical business LAN
manager will probably never see the problems I do, because most
typically used LAN protocol suites stop working long before you get to
the size the Internet is now. Conversely, this is why I have such
confidence in advocating TCP/IP for all "enterprise-wide" networking
needs in our shop: because there is an existence proof that the
technology will scale beyond the largest size Apple Computer could ever
hope to attain: The Internet itself.
OK, so IP isn't fully "plug & play" now. There are people working on
that problem. For those who would cite AppleTalk's success in this
area, it is important to note that we did not remove the complexity of
networking - we just hid it in the routers.
Erik E. Fair apple!fair fair@apple.com
