Re: [nmrg] Re: [Research-funding] updates research funding draft

[Juergen Schoenwaelder <[e-mail]>]

On Mon, Oct 20, 2003 at 01:11:20PM -0700, Sally Floyd wrote:

[...]

> * The subsection on Autonomous Network Management, originally from
> draft-iab-research-funding-01.txt, was deleted from
> draft-iab-research-funding-02.txt in the interests of shortening
> the Network Management section.
 
> * The subsection on "Configuration Management", originally from
> draft-iab-research-funding-01.txt, was also deleted in the interests of
> shortening the Network Management section.
 
> * The new material in the NRMG's subsection on "Customer Network
> Management" is the last paragraph, and this material was not included.
> Some of that paragraph is more detailed than is needed for this
> document, in terms of the specific needs of customer network
> management mechanisms.

I have now looked at both versions again and I want to provide one 
more time some feedback. I have had issues with the section titled
"Application of Expert Systems to Network Management" in the past and 
some of our proposed text was written to overcome the problem. Let me 
explain...

The first paragraph in this section is about what we people in the
community typically call Customer Network Management. This paragraph 
has in my view nothing to do with expert systems per se. And this 
motived us to move the text into a subsection titled Customer Network 
Management in the proposal we submitted earlier. You have decided to 
cut this section and you have moved the text back into the expert system
section where I think it is somewhat misplaced. (Or is the intention to
say that expert systems are the way to produce meaningful explanations
for users? But then again is that not too specific?)

The second issue I have is that the second paragraph stresses expert 
systems (which is a technique) while the underlying goal is what people
call autonomous network management.  In fact, if you look at NM research 
publications, then quite some papers were written in the early 90s about 
expert systems in network management. But nothing ground breaking has
appeared in this area for quite some years now. If you look at more
recent NM research publications, you will notice that more substantial
research comes from the application of control theory concepts to NM.
Hence, I would feel much more comfortable if the text would have been
written to stress the goal (autonomous network management) and just
names some techniques to get there, among them being expert systems,
other AI techniques or classic control theory. This was the reason
for use to propose the section on "Autonomous Network Management"
which did not make it into the current revision.

To illustrate this, here is another concrete proposal for changes, 
basically to rework and split section 3.5.3.:

OLD:

3.5.3.  Application of Expert Systems to Network Management

   An open issue related to network management is helping users and
   others to identify and resolve problems in the network.  If a user
   can't access a web page, it would be useful if the user could find
   out, easily, without having to run ping and traceroute, whether the
   problem was that the web server was down, that the network was
   partitioned due to a link failure, that there was heavy congestion
   along the path, that the DNS name couldn't be resolved, that the
   firewall prohibited the access, or that some other specific event
   occurred.

   It would be useful to examine whether expert systems technology or
   artificial intelligence technology could be used to enhance network
   management systems, reducing the operational burden created by the
   network and improving scalability.

NEW:

3.5.3.  Customer Network Management

   An open issue related to network management is helping users and
   others to identify and resolve problems in the network.  If a user
   can't access a web page, it would be useful if the user could find
   out, easily, without having to run ping and traceroute, whether the
   problem was that the web server was down, that the network was
   partitioned due to a link failure, that there was heavy congestion
   along the path, that the DNS name couldn't be resolved, that the
   firewall prohibited the access, or that some other specific event
   occurred.


3.5.4.  Autonomous Network Management

   More research is needed to improve the degree of automation achieved
   by network management systems and to localize management. Autonomous
   network management might involve the application of control theory,
   artificial intelligence or expert system technologies to network
   management problems.

This new text is roughly the same number of words so this change does not
expand the network management section. Of course, I personally find it
a pity that configuration management is not mentioned in the document 
even though it practically seems to be a real big issue. But I understand 
that this document can only touch selected issues.

/js

-- 
Juergen Schoenwaelder		    International University Bremen
<http://www.eecs.iu-bremen.de/>	    P.O. Box 750 561, 28725 Bremen, Germany

_______________________________________________
Research-funding mailing list
[e-mail]
https://www1.ietf.org/mailman/listinfo/research-funding