RE: [atlarge-discuss] Help for us wanted: a concrete challenge

["Bruce Young" <Bruce@barelyadequate.info>]

Jefsey's proposal certianly has merit.  And I think some of our
more-technical members should strongly consider it as a project once we get
up on our feet.

Bruce Young
Portland, Oregon
Bruce@barelyadequate.info
http://www.barelyadequate.info
--------------------------------------------
Support democratic control of the Internet!
Go to http://www.icannatlarge.com and Join ICANN At Large!




-----Original Message-----
From: J-F C. (Jefsey) Morfin [mailto:jefsey@club-internet.fr]
Sent: Thursday, August 01, 2002 1:42 PM
To: atlarge-discuss@lists.fitug.de
Subject: [atlarge-discuss] Help for us wanted: a concrete challenge


Hi! @large folks,

there are two ways to make the @large important. To develop actions such as
icannatlarge.com and to developing the network itself, in the best interest
of all the end users. We are matching the first challenge - are we able to
match the second one too?

One of the key issues of the Internet and the key issue of the ICANN is the
DNS. One can dispute over it, one can also work on it and to develop it.
This is the target of the dot-root project ( http://dot-root.com - in
French right now). This project consists in organizing and testing a
multiple parallel asynchronous experimental root servers system and a root
server operations test team.

The initial team is French speaking (France, Belgium, French speaking
universities people), but we want it to be something for everyone. In order
to better obtain funding we plan three test parallel root systems. From our
initial contacts one could be European, one from the Pacific area and one
International. Ideally we would be better off with five systems (one by
@large region?) for security/stability reasons.

The target is that the root systems may monitor each other for service
consistency: would an alarm occur, both the reporting and the reported root
serve system would be frozen and investigated. Part of the experience is
obviously to discuss, implement, test, validate and document what this may
means, if it is feasible, what are the cons and pros, etc.

This is absolutely not alt(sic)root, however it will certainly use
experience and tools from everywhere. If falls under the ICP-3 testing and
experiments provisions.

The target is to gather operational competences, explore real needs in
terms of parallel master file generation, updates, security, compliance
with local requirements, mutual consistency permanent control, security by
parallel queries, procedures in case of emergency, statistics, mutual
comparisons of operations, comparisons between different DNS programs,
etc.. also to look at the best way to inform the operators in case of
crisis and pick load (war, catastrophes, hertz disruptions, ..). Develop
procedures.Type of hardwares, network locations, real life situations
depending on the load. Simulation of medium, large, huge numbers of TLDs.
The way to support new architectures and different requirements in term of
root information dissemination. Three of the first projects are:

1) to build a test program to emulate the requests to such a root server
system.
2) to develop an e-learning system on the DNS for root system operators,
ISP, TLD Managers, DN managers, extranet and Intranet managers, as
different paths, so they may have a dedicated tutorial and e-assistant. We
will add management assistance tools.
3) a generic TLD management system. A few TLD test name servers will be
deployed to support root testing. One will be dedicated to a testing among
Telecom operators (dot-telecom).

All what will be done and developed will be reported and under GPL. This
experimental system will be maintained in proper order to serve as an
emergency back-up for the root system or for remote ccTLDs.

Right now we are looking for the following help:

1. Universities, corporations or experts interested in sharing into the
experiment in dedicating one Pentium+ at a stable IP address and in
participating with a permanent one or two persons  team. This will be at
least for 6 months, but probably on a much longer term. They should be
prepared to join forces to develop test/development projects and to file
for R&D common budgets. Dedication, but no special skill required -
dedicated individuals are welcome, but this is a serious and long term
project (your ISP may be interested in participating with a machine and
having his name quoted). The idea is also to test the training from scratch.

2. the site right now is in French: we would be glad if someone could help
in translating the document in perfect English, Spanish and may be other
languages so it can be presented to the press and explained to the public.
The Internet is our common network. We want to ask everyone their needs
about the DNS evolution.

3. this kind of project is going to require some non DNS support such as:

- legal writing for consortium agreement, licenses, etc also one of the
target is to be able to advise Govs on the way to deal with DNS and root
issues in the laws (archiving, testing, validity of the proof that may
represent the record of an access, disputes between ISPs and root service
providers, emergency situations, insurance contacts ...)

- local press relations to keep the press informed in case of interest.

- hackers. One of the target is to work on security. A test root system
will permit impossible testing on the main system. We will also want to
know the cons/pros of parallel solutions.

- money collectors. If anyone as skills and experience in promoting this
king of R&D projects.

I underline that this is a real occasion for the @large to show that we are
the network, that we can contribute to our own R&D and operations, that we
are mastered by no one

Also a real occasion for every network community to share. Ideally we would
need teams and machines from every country to build a common command of key
issues (DNS evolution) and to get statistics and requirement from every
part of the network, to best compare. In particular the iDNs issues are of
major importance for the Minc, Ainc and Eurolinc areas. To build a basis of
experience in these countries can only help our ccTLDs, Governments and
local development.

Thank you for your cooperation.
jfc





---------------------------------------------------------------------
To unsubscribe, e-mail: atlarge-discuss-unsubscribe@lists.fitug.de
For additional commands, e-mail: atlarge-discuss-help@lists.fitug.de