Some thoughts on the "Restructuring of RIPE"
- Date: Mon, 9 May 1994 14:22:08 +0200
- Date: Wed, 04 May 1994 22:47:34 +0200
- Address: Kruislaan 409, P.O. Box 41882, 1009 DB Amsterdam, the Netherlands
- Organisation: Nikhef-H (National Institute for Nuclear and High-Energy Physics)
- Phone: +31 20 5925102, +31 20 6924218 (home)
- Resent-date: Mon, 9 May 1994 14:22:08 +0200
- Resent-message-id: <9405091222.OA23172@localhost>
- Telefax: +31 20 5925155
- Telex: 10262 hef nl
At the last RIPE meeting we had a BOF on the structure of RIPE. It was
stated there that further discussions would be held over the mailing
list. Up till now this subject has been very quiet and the next meeting
is in 1 1/2 week. Therefore I would like to share with you some of my
personal thoughts on this subject. Having attended the last IEPG and
IETF meeting in Seattle I saw some very nice parrallels with the
discussion at the RIPE BOF. I will come to these in thes message.
In the oldest RIPE document, ripe-001, "Terms of reference", I see, i.a.
the following bullets:
- RIPE acts as a forum for the exchange of technical information
and the creation of expertise on IP networking.
- RIPE promotes and coordinates interconnection of IP networks
within Europe and to other continents.
- RIPE establishes agreement on common network management practices
and the operational management of the interconnected networks.
- RIPE serves as a focal point for other common activities of the
participants related to IP networking.
This document is dated "29 November 1989", so it is 3 1/2 years old,
which is ofcourse very, very old in networking land. Never the less
these points, especially the first and the last one, seem to me as
valid as they were at the time of writing.
If I compare this with the recently established IEPG charter, I see a
Internet Engineering and Planning Group
Domain of Activity
The Internet Engineering and Planning Group (IEPG) is a group principally
comprised of Internet Service Operators.
The common objective of the group is to promote a technically coordinated
operational environment of the Global Internet.
The activities of the IEPG will be within the following areas:
- To facilitate the operations and management of Global Internet
- To promote the introduction of new Internet services within the
Global Internet, and
- Liaison with related Internet Operations groups and liaison with
technical development groups.
Both documents talk about operations and coordination. Also I think the
"feeling" behind both is about the same. This gets me to my first point:
RIPE as a body still has the same reason to exist as it had when it
started. Perhaps some wording needs to be changed, removing explicit
reference to IP.
To me the two focal points in the terms of reference and in the work that
is being done within RIPE are "coordination" and "information exchange".
Another absolutely key word is "technical". RIPE (and I think IEPG and
other bodies with about the same goal) are technical fora and not meant
for political discussions. This does not preclude "politicians" to attend
RIPE meetings and being active in them. I think it is very worthwile
they do, as RIPE is the forum to discuss the technical implications and
(im)possibilities of the things the politicians come up with. But these
discussions should be on a technical level and open minded.
Coming back to the focal points I mentioned, "coordination" and
"information exchange", I think I am right in presuming "information
exchange" is a sine qua non for "coordination". Since RIPE started the
European Internet has changed quite dramatically, in the sense that in
the old days we had more or less a bunch of international lines owned by
the two parties at both ends of the line (some times in collaboration
with more parties), that were - slightly exagerated - randomly put all
over Europe. Nowadays there are a number of European Internet Service
Providers that have brought at least some structure in the chaos. The
point I want to make is that the level or type of coordination related
to getting a Pan-European Internet Structure has also become different.
This means that we'll see a more compex level of interaction at RIPE
meetings because we see two types af attendants, one from the "national"
or "regional" level, and one from the "european service provider" level.
It is the service providers who have to do most of the coordination and
ofcourse RIPE is the perfect forum for this because it is here that all
the best and brightest in (at least) European networking are assembled.
Let me try to define some areas where coordination may be needed. This
list is by no means meant to be complete. Some of these items are taken
from the Opera BOF at the Seattle IETF (BTW I am very fond of opera!)
- routing registries
- route servers
- exchange points
- transatlantic connectivity
- Central and Eastern European developments
- Address allocation
- "virtual" networks/systems
- Information systems like Gopher, WWW
- Directory services lik DNS, X.500
- Security issues related to IP
- Security issues related to routing
- network management
- trouble ticket hand-off
The above mentioned areas could be the base for the structure of RIPE. I
think "area" is a better wording for these groups than "working group",
because the main goal is coordination and information exchange and not
so much "work" in the sense that there are clearly defined deliverables.
Of course within a area some documents could be needed and for the
writing of such a document a working group could be formed. Also
other reasons for working groups can be thought off. The point is that
working groups are supposed to have a limited TTL and a clear
deliverable, where areas are much longer living creatures. At the RIPE
restructuring BOF the need was mentioned for a mechanism to start
working groups. I think this is indeed needed, and would have to be
defined both for areas and working groups in the sense I have defined
Some of the areas mentoned above already have a working group in the
current sense, like routing. Infrastructure could be the subject of the
EEPG working group.
Where information exchange is at the one hand needed in order to be able
to coordinate on the other hand it also is a topic of its own.
One of my goals in attending RIPE meetings is to learn things. And one
of the means is attending presentations, either by the RIPE NCC people
about their wonderfull new tools or by people who want to tell about the
things they are doing. This can be on almost any level, f.i.
the presentations on the status of Ebone or EUnet or Dante, or the talk
Brian Carpenter gave last meeting on IPng.
This "information exchange" or "learning" topic also relates to the
issue raised at the last RIPE meeting (and one that was also raised at
the open plenary at the last IETF) about the size of the meeting (both
plenary and working groups). Bigger amounts of people will change the
character of a meeting and will get a smaller percentage of the people
who are active and participating in the discussion. I do not see this a
s a problem, because by listening to discussions that have a fair amount
of openness, you can learn quite a lot. Having a discussion by
knowledgable people is a way to disseminate this knowledge to relative
newcomers. So I should say, the bigger the RIPE meetings are, the better
RIPE can fulfill its goals.
I hope the points mentioned above can be used to help the discussion
about the goal and structure of RIPE along a little.