Message-Id: <6.2.1.2.2.20050812101801.050755a0@mail.jefsey.com>
Date: Fri, 12 Aug 2005 11:07:12 +0200
To: Bill Sommerfeld <sommerfeld@sun.com>, Stephen Kent <kent@bbn.com>
From: "JFC (Jefsey) Morfin" <jefsey@jefsey.com>
In-Reply-To: <1123798200.14955.137.camel@thunk>
References: <42FA5884.9060103@cisco.com> <42FB1640.5020906@zurich.ibm.com>
	<42FB664A.60605@cisco.com> <01LRP7VL7OTW000092@mauve.mrochek.com>
	<p0623090cbf21597402ad@[128.89.89.106]>
	<1123798200.14955.137.camel@thunk>
Mime-Version: 1.0
Content-Type: text/plain; charset="us-ascii"; format=flowed
Cc: Ned Freed <ned.freed@mrochek.com>,
	Michael Thomas <thomasm@cisco.com>, ietf@ietf.org
Subject: Re: what is a threat analysis?
Precedence: list
Sender: ietf-bounces@ietf.org
Errors-To: ietf-bounces@ietf.org

At 00:10 12/08/2005, Bill Sommerfeld wrote:
>So an effort to come up with a consensus threat analysis sounds like a
>very good idea.  It might even be worthy of a working group of its own
>as it would likely be useful as a base for more than just the MASS/DKIM
>work.

Agree. After 9/11 ICANN changed all its priorities towards security. I 
proposed the BC (Business Constituency) to work on an "ICP-4 Network 
Security" proposition (ICPs are the ICANN documents). Several large 
operators, manufacturers contributed to the project which went to an end 
rather quickly for Layer 8/9 reasons. The first things we tried to do was:

- to list all the possible types of threads on the conceptual network (this 
does not include the user security by itself).
- all the motivations for attacks - from war down to employee retaliation 
or flooding, etc.

We saw it was an iterative process. Motivations helping identifying 
threats, and vice versa. I am not sure I still have a copy of the work we 
carried (it was privately made in the paranoia of security which was the 
mood), but I established a method to help it. Anyway from this experience I 
would suggest the Charter of such a group should be discussed with an 
enlarged committee including Steve Crocker's ICANN security committee, 
people from Govs, Military and Commercial Intelligence, people from GAC, 
people from targeted industries (Banks, Airlines, critical infrastructures, 
life depending solutions).

I used this experience when considering the network security aspects in the 
dot-root test-bed, carried in 2002/2003, and for the resulting "national 
vulnerability to the internet" meetings I organised for France in 
2003/2004. This lead to the identification of the need of a 
compartmentalisation of the internet for risk containment, intelligence 
protection, community development, service innovation, spaces of exhanges, 
trust and services, etc. and of the need to analyse, organise and run its 
intergovernance. We incorporated the AFRAC as a national internet community 
effort to work and test on the matter.

We observed that it was the best way to achieve the necessary virtual 
partitionning of the internet while preventing the grassroots balkanisation 
which develops. Basically it means that the unicity and the unity of the 
global network will be supported and stenghtened by a myriad of well 
established, suited and protected externets (virtual external network 
look-alike). Security of a distributed network is to be distributed.

Also, one has to consider the external implications of the security 
failures and understand that threats can be to second, third etc. degree, 
the implication can be major and concern (due to the today implication of 
the Internet in the world's life) the life of millions. This was the first 
study, which served as basis for the thinking, and the most important one 
todate due to its implications: http://whitehouse.gov/pcipb and its 
preparation is certainly a basic input to consider in that area. Question 
is: is that IETF? IETF is probably more interested in the resulting 
specifications? The implications on the current internet structure and 
vision are important.

jfc



_______________________________________________
Ietf mailing list
Ietf@ietf.org
https://www1.ietf.org/mailman/listinfo/ietf