[470] in Commercialization & Privatization of the Internet
Re: ANS Acceptable Use Policy
daemon@ATHENA.MIT.EDU (Joe Ragland)
Fri Mar 29 01:38:35 1991
Date: Fri, 29 Mar 91 01:37:02 -0500
From: Joe Ragland <jrr@concert.net>
To: emv@ox.com
Cc: com-priv@psi.com
> I am curious as to what these serious problems were (or are), when they
> occurred, and why they need to be brought up in your policy statement?
Guess you weren't around an NOC back before Sun fixed nameservice via YP
when local YP master servers could gang up on some remote Internet server
(usually down) and generate domain requests to the remote network at
full circuit speeds (T1). Trouble tickets were not sufficiently fast to
get to the source but the NIC whois database and the phone usually did the
job. The destination net thought this event to be immensely damaging
by what I heard said on the phone a time or two.
The CONCERT policy was written to cover such situations and to handle
a mostly bridged environment that existed many years. Then, much of
"what appeared on the local ethernet" got replicated on many other CONCERT
ethernets and could do immense dammage. Like mis-configured proxy arping
Kinetics FastPaths which can redirect all of your network's routing to
the K-box. We consider such an event immensely damaging. The policy is
aimed at getting local admins to think a bit before they take it out of
the box and plug it in to see if it plays...
As to the policy, we welcome any comments. Somehow, I think the subject
has changed but that's fine too.
Joe Ragland
