RIPE NCC Review - Draft
-
To: ripe-list@localhost
-
From: k13@localhost (Rob Blokzijl)
-
Date: Tue, 14 Sep 1993 12:44:15 +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)
-
Telefax: +31 20 5925155
-
Telex: 10262 hef nl
DRAFT DRAFT DRAFT DRAFT DRAFT DRAFT DRAFT DRAFT DRAFT DRAFT DRAFT DRAFT DRAFT
RIPE NCC Review
R.Blokzijl
M.Sterba
W.Woeber
7 September 1993
Version 1
1. Background
The current activities of the RIPE NCC are guided by, and based upon,
the document "RIPE NCC Activity Plan" [ref.: ripe-035]. This document
was accepted by the RIPE community in May 1991 as the bases for a
future RIPE NCC.
Since April 1992 the RIPE NCC has been operational on the basis of the
set of activities as described in ripe-035.
At the RIPE meeting of April 1993 it was decided that a review of the
first year of operations would be a needed in order to determine
whether
a. the goals as put down in the Activity Plan have been met
b. an update of said activity plan is needed
This document gives a review of current operations of the RIPE NCC. An
updated activity plan will be presented in a separate document.
The document is structured in two parts. The first part gives a
detailed evaluation of the individual activities, as defined
in the Activity Plan. The second part contains remarks of a more
general nature.
2. Evaluation of Technical Activities
2.1 Technical -- Coordination Activities
2.1.1 The RIPE Database
This task is performed by the NCC according to the Activity plan.
New objects and attributes have been added as agreed by the RIPE
community. New tools for checking, maintaining and distributing
the database are being implemented.
Reports are handled through the Quarterly Reports in an excellent
way.
2.1.2 Distribution of the Database
The distribution of the database is handled well. The newly
developed software will support consistency between primary and
secondary copies better.
2.1.3 Coordination of Database exchange with other Organisations
This is permanently pursued. In cooperation with the Internic and
Merit an exchange format has been defined. Good progress has
been made, the delay in implementation of the actual exchanges is
not attributable to the RIPE NCC.
2.1.4 Keep a record of IP connectivity in Europe
This is progressing, albeit slowly. However we think this activity
could be reviewed in the light of recent developments on the
organisational background of networking in Europe. A more
distributed approach could be advisable.
2.1.5 Keep an inventory of IP infrastructures not yet interconnected
Basically this is done via the RIPE DB, though this could be only
up to date from the moment the NCC started registering IP
networks for Europe.
2.1.6 Keep a record of operational contact points
This has been implemented by way of mailing distribution lists
that enable subscribed networks to keep in contact.
2.1.7 DNS Software
Operational issues are coordinated by the RIPE DNS WG. The NCC
maintains an archive of relevant software and documentation.
2.1.8 Placement of name servers and backup servers in Europe, and
beyond
From our experience, this currently appears to be a non-issue.
At least down to the level where central coordination and
checking can be done, there seem to be no major problems that
would call for allocation of NCC manpower.
2.1.9 Gather traffic statistics and network performance data
Basically no progress has been made in these fields. It is our
opinion that these activities belong to various network
operators, not a central organisation such as the RIPE NCC.
The RIPE NCC could provide a store facility for those providers
who wish to publish their statistical data on neutral ground.
However, the role of the RIPE NCC in gathering quality of service
data needs to be discussed by RIPE.
2.2 Technical -- Development Activities
2.2.1 DB Management Tools
The development of more sophisticated DB management tools has
been an ongoing activity, resulting in a very stable and able
database service. The definition of new tools has been done in
close cooperation with the RIPE Working groups on Routing,
Local-IR and DB, thus ensuring the fulfillment of user needs.
2.2.2 DB Querying Tools
This has resulted in improved versions of client software that
are being used widely. Also the integration with most commonly
used information retrieval tools is excellent.
2.2.3 Network Statistics and Performance gathering and presentation
tools
Not much progress has been achieved. The need for these tools is
clear, but a more precise definition of the role, and the amount
of resources, of the RIPE NCC is needed.
2.2.4 DNS Quality Control Tools
Tools are available from the RIPE NCC server. The NCC itself does
a monthly analysis of the European DNS space, resulting in the
famous hostcounts, and a set of comprehensive error reports for
all European domains. It is our opinion that further quality
control of the DNS space, is up to the responsible organisations
on a top level basis.
2.2.5 Tools for drawing network diagrams
Basically no progress has been made in this area. A repository
and a common naming scheme for network maps have been implemented.
We do feel that the work should have some priority, but whether
it should be done by the RIPE NCC is debatable. A special project
might be more appropriate.
2.3 Administrative -- RIPE Activity Support
2.3.1 Keeping the RIPE document store
This has been implemented in an extremely friendly way for the end
users: all commonly available information retrieval tools give
access to the document store. The documentstore itself is kept up
to date, and has been implemented in a well structured way.
2.3.2 Technical and secretarial support for studies undertaken by RIPE
This has been working quite well. We should probably try to put
in place a chasing mechanism for work items issued from RIPE
meetings i.e. more follow-up on the WG's and WG chairs.
2.3.3 Organisational support for RIPE meetings
This works extremely well, and has substantially augmented the
productivity of the RIPE meetings.
2.4 Administrative - Reporting
2.4.1 Quarterly reports on all activities will be produced for the RIPE
participating organisations and the RARE CoA
2.4.2 Quarterly reports on outstanding problems will be produced for
the RIPE participating organisations and the RARE CoA
We conclude that excellent quarterly reports have been produced
in a timely fashion. They consist of detailed descriptions of the
accomplishments of the RIPE NCC, a discussion of outstanding
problems, and excellent overviews of the state of the Internet in
Europe. They not only serve as administrative reportings, but
also as reference material for network administrators and
operators.
2.4.3 the NCC reports on its ongoing work and outstanding problems
also to each RIPE meeting.
This has been done at each RIPE meeting. Especially presentation
of outstanding problems is of importance, since in most cases
participation and decision by RIPE is needed to resolve them.
2.4.4 The NCC will produce annual reports
At the time of writing of this report, the first RIPE NCC annual
report has just been produced.
3. General comments
We have concluded that the RIPE NCC has in its first year of operations
achieved the goals that were aimed at back in 1991. The flying start of
most of the activities within the first few weeks of operations has
been extremely impressive. Even more impressive has been the way the
RIPE NCC staff has been able to keep up the momentum.
A careful analysis of the original Activity Plan has shown that in a
few areas the progress has not been as was foreseen. We recommend that
RIPE has a careful look at these areas and decides on their relative
importance for the near future.
On the other hand, we have found that certain areas of activity have
taken much more resources then was originally foreseen. The activities
concern mainly the IP Registry function of the RIPE NCC. It is
interesting to notice that this activity is not even explicitly
mentioned in the Activity Plan!
Another area that the current activity plan has not foreseen is the
involvement of the RIPE NCC in special projects. This requires
resources on the management level and on the technical level. Though
the projects are formally separately staffed and funded, the fact that
they are housed in the RIPE NCC automatically means that some NCC
resources go into these projects. This has been intentional, but it is
not covered by the Activity Plan.
We conclude that the RIPE NCC has become a focal point for development
activities in many areas of Internet technology. We therefore strongly
recommend that RIPE defines a new Activity Plan for the RIPE NCC, taking
into account the analysis above, and the actual set of activities of the
RIPE NCC as they have developed over the first year of operations.
4. Acknowledgements
We would like to thank the following persons for their involvement in
preparing this report:
Daniel Karrenberg dfk@localhost
Tony Bates tony@localhost
Anne Lord anne@localhost
Marten Terpstra marten@localhost
Daniele Bovio hi@localhost
Nandor Horvath horvath@localhost
Francis Dupont <francis.dupont@localhost
Jean-Michel Jouanigot jimi@localhost
Bernhard Stockman boss@localhost
Blasco Bonito bonito@localhost
Vladimir Vrabec vrabec@localhost
Geza Turchanyi turchanyi@localhost
Ruediger Volk rv@localhost
Roland Acra acra@localhost