RIPE 15 Minutes

1. Opening

1.1. Welcome

Rob Blokzijl welcomed the participants to the 15th RIPE meeting hosted by NIKHEF and held in Amsterdam.

1.2. Approval of the agenda

The agenda was approved.

1.3. Papers tabled

Agenda for the 15th RIPE meeting
RIPE Quarterly Report, Issue 4, March 1993
The RIPE NCC "One year of experience with European Internet Coordination" - DRAFT
Working Group Agendas
Map of International IP connectivity in Central & Eastern Europe
RIPE NCC Funding Paper (Doc ID: ripe-84)
How to make better use of the existing infrastructure?
Draft NCC funding letter

Presentation notes

IXF An Interconnect model for Internet Service Providers and Network Operators
IETF Report
NASA: Russia connectivity

2. Minutes of the last meeting

2.1. Approval of the minutes

The minutes from the 14th RIPE meeting were approved. The action items from the 14th RIPE meeting minutes were reviewed. The following list comprises the ongoing action items only. All other action items were closed.

Action: 13.3 Marten Terpstra
Changes in database format. Coordination discussions with US counterparts are ongoing.

Action: 13.4 Marten Terpstra
To progress GSI/Merit/RIPE NCC exchange format discussions.

Action: 13.6 NCC
To include the operational contact attribute in all object definitions besides persons.

Action: 14.4 NCC
Generation of Internet handles by local registries. The InterNIC is working on it.

Action: 14.17 Milan Sterba
Track the US development of the resource-object description work and will report to the list (see also action 15.9).

3. NCC Funding Arrangement

3.1. Status report (R Blokzijl)

A paper has been circulated (ref. ripe-84) which proposes a future funding model for the RIPE NCC. The document will form the basis for future RIPE NCC funding discussions. The current model proposes a "how much" model which defines a tariff relating to the size of an organisation. "Who should pay" has been based on charging Internet service providers since they are the direct beneficiaries of the services of the RIPE NCC.

To progress this activity, a draft letter is currently being approved and will be circulated once finalised (authors: Rob Blokzijl and Paul Van Binst) together with the funding paper (ripe-84) and other RIPE NCC publications to solicit funds from potential contributors for 1993 and for committment following the proposed guidelines.

3.2. News from RARE (T Kalin)

To secure funding for 1994, invoices to all current RARE members will be sent out shortly. Contributions from RARE members will secure the majority of funds required for RIPE NCC for 1994.

At the forthcoming CoA in Trondheim, RARE will formally offer to continue the existing arrangements for RIPE i.e. RARE is to provide the formal framework for RIPE as already established. The RIPE community confirmed their acceptance of this offer and Tomaz Kalin
confirmed that this acceptance would be conveyed to the RARE CoA.

It was suggested and agreed that the chairman of RIPE should be given mandate to try to arrange for a long term solution to this problem. To summarise: long-term stability for the RIPE NCC must be based on a proper and secure financial scheme. One such scheme has been outlined in the recent funding paper (doc ref. ripe-84) and the participants of
the meeting were urged to further discuss this in the coming months as a matter of priority.

4. Working Group Agendas

The chairmen of the respective working groups briefly introduced their agendas and the work items to be covered.

5. RIPE NCC Review Board (R Blokzijl)

Rob Blokzijl briefly described the history of RIPE NCC and origin of the activity plan. The RIPE NCC is now one year old and a review of the activity plan is necessary to establish whether the needs of the RIPE community are adequately covered.

Three deliverables were identified as necessary:

evaluation of the operation of the RIPE NCC project 1 year on - timescale: 2-3 weeks
evaluation of the activity plan - timescale: next RIPE meeting
wish list document of revised RIPE NCC activities - timescale: next RIPE meeting

A number of members of the audience echoed the sentiment that the neutral role of the RIPE NCC must be clearly defined and maintained. It was also suggested that the accountability of the RIPE NCC could be considered.

6. RIPE NCC Report (Daniel Karrenberg)

The RIPE NCC Report was based on the Annual Report: "The RIPE NCC One Year of Experience with European Internet Coordination" (Karrenberg, Lord and Terpstra) which was distributed as a tabled paper. The report will shortly become a ripe document retrievable from ftp.ripe.net. An announcement to the RIPE mailing list will be made when it is available.

7. NASA Science Internet extensions into Russia (Allyson Showalter)

A presentation was given on the "The NASA Science Internet (NSI)" and the plans for "Extensions into Russia". The presentation covered topics such as the NSI Charter, Acceptable Use Guidelines, the NSI Approach, NSI Science Requirements for Russia, IKI as NSI "Gateway" to Russia and the schedule for Russian connectivity. Maps were also displayed showing the Russian Space Science Internet (RSSI), IKI: NSI Gateway and Bitnet
Hub and maps of the NASA Science Internet.

Hard copies of the presentation were distributed at the meeting and are available from the RIPE NCC on request.

8. An Interconnect model for Internet Service Providers (P Lothberg)

Peter Lothberg presented his ideas concerning a possible method of scaling the Internet in the future with multiple providers. Hard copies of the presentation were circulated at the meeting. The slides are available on request from him.

9. How can we make better use of the existing infrastructure? (P Jones)

Some ideas for possible improvements in the use of the existing Internet infrastructure were presented including the following topics for consideration

Possible topics (A):

white pages - contact info on persons
email
pictures
voice
output to fax
Time services
Interaction: in general, x.29

Possible topics (B):

IP and multiprotocol working
x.25
CLNS
DECNET
OSI Applications
CONS
(Others - Earn, Proprietary)

Methods:

exchange info
catalogue references
investigations

Mechanisms:

RIPE discussions and mailing lists

Summary:

how can the Internet be exploited further?
is there a way to go forward?

Commenting from the chair, R Blokzijl stated that the RIPE charter specifies that the community is restricted to coordination of tcp/ip networking, and pointed to the working groups within IETF/RARE who are tackling these issues. However, it was acknowledged that the growth of the Internet is such that locating information is often not a straightforward process (as the information on the Internet continues to grow rapidly) and much work is still to be done. The RIPE community however, should watch to see where relevant co-ordination at the service provider level may be necessary and act when it is needed.

10. IETF Report (E Huizer)

Erik Huizer gave an overview of the structure and work of the IETF and urged members of the audience to attend and participate in the forthcoming IETF conference to be held on July 12th-16th in Amsterdam. This IETF is the first to be held in Europe.

11. Delegation of zones in the 193.in-addr.arpa domain

The idea

193.* delegated to the NCC
193.x delegated to local-iI
Hierarchy in in-addr domain created

Procedures

x.193.in-addr.arpa delegated to local-ir that administers 193.x if local-ir wishes
2 servers for each x.193.in-addr.arpa zone AND secondary on ns.ripe.net
x.193.in-addr.arpa servers must have connectivity like top level domain server
delegation does NOT mean control, it means administration
reverse zones inside x.193 when x.193 is not delegated is done by NCC, plan to use rev-srv field in RIPE database to update automatically

Changing Provider

service provider can NOT force changing leaving customer to renumber
y.x.193.in-addr.arpa service must be continued, even though customer is no longer yours!

How to request

Send in domain object for x.193.in-addr.arpa to <[email protected]> as shown in the example below:

domain: 202.193.in-addr.arpa
descr: Pan European Organisations block
admin-c: Daniel Karrenberg
tech-c: Marten Terpstra
zone-c: Marten Terpstra
nserver: ns.eu.net
nserver: sunic.sunet.se
nserver: ns.ripe.net
changed: [email protected] 930319
source: RIPE

The NCC will check, delegate, run secondary and forward entry to RIPE DB.

Current Status

193.in-addr.arpa delegated March 15 by NIC
block delegation started end of March
21 blocks delegated in 193.in-addr.arpa
ns.ripe.net running secondary for all as specified in procedures
single net delegations in 24 other blocks
procedures are draft version 1.4, will be RIPE document after this week
no problems discovered so far
Thanks to Havard Eidnes for the idea!

Slides of the presentation can be retrieved from ftp ftp.ripe.net in directory: ripe/presentations/ripe-m15-marten-IN-ADDR.ps.Z

12. RIPE / RARE Joint projects progress report

12.1. Route Server Joint Project Progress Report

Three main components for the project

The physical "GIX" itself
The "Route Server (RS)"
The "Routing Registry (RR)"

Main Goals

Establishing a route server on the GIX itself
RS aspects as opposed to transport
Making it possible to document European routing policies to make RR viable

Routing Registry

Neutral run registry (RIPE NCC)
Would manage routes for a region or domain (model used - Geographic)
Registration of routing policy information
Proposal for a format for documenting routing policy "Representation of IP Routing Policies in the RIPE Database"(ripe-81)
Generic guidance document as well as having direct relevance to European Route Server work

Routing Registry

Database and objects must be easily derived for use by the route server. Tools for information and debugging are currently being developed for routing policy

"prcheck" - consistency checker
"prpath" - extract AS-path information
"prconn" - display connectivity
"prtrace" - policy derived traceroute
Configuration tools
Network-based lists generator
Interactive Configurator
RS/RR interaction tools

Plan

12 point plan produced (work in progress) - summarised below: (refer to paper for timescales)

Routing Registry

New database objects in RIPE database
Modification of software
Impact document for providers
Development of database software
Development of Tools
Development of update and transition procedures

Route Server

Install Route-Server on the Proto-GIX itself
Test and possibly enhance route server software (keep focus of initial RS ideas of ripe-82)
Work in collaboration with the NA route server and any future route servers
Clear and smooth transition with the providers at the GIX
Use the RS for operational traffic (ULTIMATE GOAL)

Status - Routing Registry

Ripe-81 produced and general consensus seems very good
Large effort of data collection underway
Positive results so far (some way to go yet)
"25% of ASs have policy information"
Daily generated information of current routing policy information as seen "in routing tables" as opposed to the RIPE database can be fetched from ftp.ripe.net:ripe/as
Database software enhanced (whois for AS info)
Prototype configuration tools working for cisco and gated network and as-path based filter lists
Prototype "prtrace" tools working

Status - Route Server

Route Server installed at Sprint POP - thanks to Sprint, SUNET and AlterNET
SUN IPC running "gated" alpha
Initial testing done using some test-networks in Amsterdam and Stockholm
Currently peering with ENSS and ICM-DC-1 as a test doing third party routing correctly
Generated European list could be used today

Status

Development RS installed in Amsterdam on IBR LAN (Ebone Amsterdam DMZ)
Test cisco installed (thanks to SURFnet)

General

Good response from Providers at the GIX
On Schedule with proposed time-scales
Going Well!
CIX association will use tools and ripe-81 for their own Route Server on the GIX

Slides of the presentation are available from ftp ftp.ripe.net in directory ripe/presentations/ripe-m15-tony-RS-SUMM.ps.Z

12.2. GISS

What is Giss?

RARE project as part of the RARE Technical Program being carried out at the RIPE NCC
Project funded by SURFnet through RARE
Open project
No Hidden agendas
Should not be seen in any way as a competitive / political document
Not just a paper exercise
Should be worthwhile and useful for new and old in the internet

Status (1)

Prague RIPE, BoF
Started discussion and working on focus
Creation of Working Group: "[email protected]"
To Join: send mail to: [email protected]
"Strawman" proposal produced and distributed
Workplan produced
Basic idea is evolvement by working group ... so input is badly needed.

Status (2)

Columbus IETF held a GISS BoF
Very good US reviews
Created a list of aspects to be detailed

Focus

Very much at the provider level rather than at user level
RIPE Amsterdam
First working group meeting
Second Strawman proposal is the outcome of this work (conscience)
Focus and scope is decided or at least for the short-term
Clearly needs to be a long-term project..

Status

Hand off areas to other groups specifically the Connectivity Group
Try to widen scope of connectivity group making use of the GISS aspects
NOT EASY TO GET BOOT STRAPPED
happy with scope and structure of project

Slides of the presentation can be fetched from ftp ftp.ripe.net in directory: /ripe/presentations/ripe-m15-tony-GISS-SUMM.ps.Z

13. RIPE Routing Policy (R Blokzijl)

The document "Representation of IP Routing Policy in the RIPE database" by Bates, Jouanigot, Karrenberg, Lothberg and Terpstra (Doc ID:ripe-81) was accepted as a ripe document. Work on an earlier version of these ideas (Doc ID: ripe-60) is no longer recommended. Those using the earlier style of registering routing policies should contact NCC and migrate to new method of registering their routing policies. Extensions of ripe-60 will not be accepted as from April 29th and migration should take place as soonas possible.

Action: 15.1 NCC
Implement update procedure with respect to ripe-81

14. Short reports on developments in European networking

14.1. RTC report - Milan Sterba

Two meetings of RTC subsequent to last RIPE meeting where the following topics were discussed:

the planning of the 15th JENC conference in Trondheim
effectiveness of RARE Technical Process: WG's, TF's, RTC, RTR's
IETF relationship: RTC planned to become peer to IETF for voice in Europe
changes in the financial scheme: meetings of working groups are currently financed by RARE but a changing the arrangements is currently a discussion item.
RTR's reviewing process
RARE Technical Program: definition of the overall strategy needs to be clarified
Current Task forces: CLNS, ISUS, CHAR and proposed task forces: ATM, SECURITY
IP next generation pilots: invitation to RIPE community to join pilots.

14.2. Developments of EuropaNET (S Moller Nielsen)

EMPB - Data Circuits:

8Mbps trunk circuits

IP pilot: as of march 1st, 1993 4 regional networks connected

SWITCH running EGP - native on 64 k lapb, x25 on 128 kbps
JANET running BGP over x25 on 2Mbps
GARR BGP over x25 on 64 kbps encapsulation techniques
RCCN BGP over x25 on 64 kbps

IP host in EuropaNET for DNS, ftp and connectivity tests.

Status CLNP

Access:

HDLC (LSPB, iso 8880-3)
x25 (iso 8473, clause 8.4.3)
FR (core x.992)

Routing:

Static
IDRP

CNLP Pilot (started march 1) - 4 regional networks are connected

switch
funet
surfnet
rediris

All using static routing/routing updated distributed via clnp pilot list.

IP and clnp pilots:

CLNP (iso 8473)
CLNP internet connectivity now
TUBA demonstration at JENC
IP internet connect during May Gateway arrangement with EBONE

15. Reports from the working group parallel sessions

15.1. Local-ir Working Group (D Karrenberg)

Review of ripe-draft-in-addr-procedures-v1.4.txt

The document was approved by the working group and will be archived as a ripe document.

Action: 15.2 Marten Terpstra
ripe-draft-in-addr-procedures-v1.4.txt to be archived as ripe document

Review of ripe-draft-hints-v1.txt

The content of the document was discussed in detail.

Ruediger Volk suggested that when address space is assigned it be made clear that the same organisation should include a report on how this address space is being used in any requests for additional address space.

Action: 15.3 Ruediger Volk.
To suggest text on the necessity for organisations to supply details of their planned address space usage when requesting additional address space.

It was agreed that the comments from the working group would be folded in, the document would be circulated via the working group list, after which the document would be published as a ripe document. The NCC offered to mirror the document in local language.

It was also suggested that the document should be put forward as an rfc.

Action: 15.4 Bob Day
To fold in the comments of the working group on the ripe-draft-hints-v1.txt, circulate the document to the wg list and publish as a ripe document.

Review of draft Delegated Internet Registry Leaflet

Minor changes to the draft document were suggested and agreed.

Action: 15.5 Anne Lord
To fold in comments to the ripe-draft-hints-v1.txt and publish the leaflet "Delegated Registry"

Review of Funding paper (ripe-84)

It was recommended and accepted that ripe-84 should reflect the RIPE position on the funding of the RIPE NCC. Major points to emerge from the discussions were as follows.

It was suggested that there should be a refinement on using the terms medium, small and large to describe organisations. Regional, national or local were suggested categories used to describe the relative size of service providers. Anyone with ideas for refining the funding categories are invited to contact NCC with their suggestions.

The problems listed in sec. 5.1 of the report concerning potential non-contributors were not considered to constitute a serious problem. It was agreed that charging individual tcp/ip users (category 4.4) should not be implemented at the present time.

It was anticipated that the eventual charges to service providers might turn out to be higher than those quoted given that more personnel may be needed at the NCC in the future.

There was some discussion over terminology: whether the term "sponsoring the RIPE NCC" should be used in favour of service providers being "billed".

It was felt by some that there must be incentives for service providers to run local registries but it was agreed that this should be discussed at a future date.

Action: 15.6 RIPE chairs, chairs of working groups and RIPE NCC manager
Meet to progress RIPE NCC funding work item

Review of RIPE NCC Internet Numbers Registration document (ripe-72)

It was agreed to recommend rather than require a 2nd opinion should be sought for requests for assignments of more than 32 C's. Some minor comments were made about ripe-72 for inclusion in the forthcoming update of the document.

Action: 15.7 Marten Terpstra
Fold in comments and republish amended ripe-72

Brainstorming

It was agreed that Mbone coordination is not a high priority work item for NCC, since it works well with current mailing list, although it was acknowledged that the situation may change.

Internet Registry workshop: plan to look at case-studies of IR experiences. Suggested to schedule for 16th RIPE meeting

Action: 15.8 NCC
To collect interesting cases of IP network number allocations (all local IR's, NCC) in preparation for the local-ir closed seminar

15.1.1. MBONE (Erik Jan Bos)

planned for Amsterdam IETF
2MB lines (GIX-STH-AMS-CERN-GIX) to be in place by May
need mrouter at EBS sites (preferred)
need coordination before and after
need tests soon
need "emergency off"

Action: 15.9 NCC
Assist SURFnet for the testing and support of MBONE for Amsterdam IETF.

AOB

It was suggested that the date of assignment of a particular piece of address space be recorded in the RIPE database. After some discussion it was agreed to propose new tags to the database working group. The tags are "created" - for all DB objects - when was this object created and "assigned" - for internet numbers and (maybe domains) when was the
number (domain) assigned.

Action: 15.10 Daniel Karrenberg
To propose new tags "created" and "assigned" to the database working group for consideration.

15.2. Database working group (Wilfried Woeber)

Status of the Database Software

The RIPE NCC is currently investigating the case for a complete rewrite of the database software. It is felt that the database has grown too large and the volume of updates is increasingly lengthening the update process.

DB Documentation

With the introduction and amendment of various objects and procedures there is an urgent need to refurbish the documentation. It was suggested also to make an attempt to remove outdated things from the DB and the documentation. Possible structure:

short list of objects, maybe along the lines of the local-ir user documentation
doc for the software and secondary whois servers

A discussion will be initiated on the DB list to propose an appropriate structure of the documentation

Action: 15.11 Wilfried Woeber
Initiate discussion on the db-wg list proposing appropriate structure of the documentation for the database

Resource Object

There was once again a request for the introduction of a sort of "Resource Object". This is a longstanding discussion and the DB group should now come to a decision soon, either to implement or abandon. There will be another attempt to synchronize with similar US activities

Action: 15.12 Milan Sterba
To research and refine proposal for the "Resource Object" tag (see also action 14.17)

Organisational matters

There was some discussion about the expectations and scheduling of the DB WG session in relation to other working group sessions. For the next meeting an attempt will be made to discuss and improve the scheduling of time-slots. Input and suggestions welcome!

15.3. Connectivity 1 (Milan Sterba)

The CEEC version 6 was discussed: orientation, structure and form of the report. Not only CEEC but all European countries will be included in future versions. Submissions will be formally reviewed before being incorporated into report and guidelines will be produced.

Action: 15.13 NCC
CEEC Report version 6 will be formatted into index data store with a hypertext tool installed which can be used to search the report.

Action: 15.14 Milan Sterba, Rob Blokzijl, Tony Bates
To coordinate and review the submissions for version 7 of the Connectivity Report.

How to get Version 6 of the CEEC report distributed before Trondheim was discussed

Action: 15.15 Milan Sterba, NCC, RARE SEC
Milan To produce version 6 of the report. NCC to print it, RARE SEC to copy and distributed by RARE SEC/NCC

The relationship of the connectivity working group to Internet monthly report was discussed. How to pursue this was considered but no action items were agreed.

The relationship of GISS to the mandate of the connectivity working group was discussed. Input from the GISS working group was felt valuable and relevant in terms of being able to provide terms and taxonomy of what connectivity services are and what connectivity is. The "metrics" of network services were discussed. It was agreed that connectivity and reachability cooperation measurements should be piloted over Europe but no actions were decided upon.

15.4. Connectivity 2 - Russia

An exhaustive list (which will be part of the next connectivity report by Milan Sterba) of planned, existing and ordered links to Moscow region was drawn up. It was noted that the internal situation regarding connectivity in Russia was a complex one. Future planning needs to focus the discussions on IP addressing and routing.

Overview:

NASA: NSI - IKI (Moscow) 256 kbps. Sites are all within Moscow region.
Department of Energy status nuclear (may use NASA link)
HEP (ITEP ) DESY 14 kbps link
ESA line IKI - Darmstadt 64 kbps
dial up DEMOS - EUnet.fi 14.4 kbps upgrade to 64 kbps and
dial up IP Freenet - NORDUnet (UNIC)
Gran Sasso to Dubna 64 Kbps line
FUNET dial up experiment
IAS SOVAM teleport (IP/NET) 9.6 kbps San Francisco - Moscow
Sprint International line from Moscow to Washington hub via GIX 64 kbps
RELCOM to Tallin via Helsinki 19.2 Kbps; not sure if operational yet.
RELCOM from relcom Moscow to St Petersburg to Helsinki 19.2 kps
Moscow State University to DESY/DFN: 64kbps
DEMOS X25 to DATEX-P 64 kbps; not used yet
Moscow to Stockholm 128 kbps line proposal

Internal coordination inside Russia is quite clearly needed. Given the above capacity and number of separate lines M Popov discussed these problems further.

Dmitry Avdeyev gave a short presentation about a joint project between DFN, DESY from German side and three HEP institutes in Moscow. A satellite line between DESY (Hamburg) and Moscow State University will be created to make access for scientific community in Russia to WIN and HEPnet. At the first stage three HEP institutes (NPI MSU, ITEP, Lebedev) in the Moscow region will be connected.

15.5. GISS working group (T Bates)

Administrivia:

Strawman 1 ideas - original ideas were reviewed
fyi 16 - input from RIPE is sought to give a European view. However no volunteers were forthcoming to undertake this work.
focus of GISS: interface between different providers

Scope:

aimed at service providers
guidance on routing and connectivity coordination
deal with provider-provider interface
stress the global aspects of the Internet
need to look at provider relevant aspects
in detail where needed i.e. dns, routing, etc
check the scope is realistic in timescales.

Connectivity overlap - Milan Sterba

There was some discussion about the need to develop a framework about connectivity in each of European countries i.e what services there are and how they should be described. Also there were discussions concerning the need to monitor connectivity quality. It was felt that a project should be initiatiated which looks at qualilty of connectivity - specifically the project would need to define tools, methods of measurement and services. No action items were agreed.

GISS document - outcome

Structure of document

provision of service rather than a specification document. Is almost a cookbook.
so it acts as a tutorial for service providers ie. do and donts, hints

Aspects of strawman 2 paper

access ip layer or gateways
generic services
connectivity
operations
information provision/co-ordination
security

3 layer document which will be focused on provider/provider relationship.

Action: 15.16
Tony Bates Publish straw 2 paper for comments and request volunteers to undertake some of the work.

A strawman 3 paper is not currently planned.

Glenn Kowack volunteered to write in collaboration with Tony Bates a top level introduction for the user as an add on to the GISS paper once it has been written.

Action: 15.17 Glenn Kowack and Tony Bates
Write a top level introduction for the user as an add-on to the GISS paper once it has been written.

15.6. DNS Working Group (Francis Dupont)

Changes in the root name server list as follows:

ns.internic.net
aos.arl.army.mil

DNS traffic: more statistics are needed for analysis

Supposed final beta version of BIND 4.9 is out

improvements in cookbook and in documentation

Next generation of IP: ftp.sesqui.net

TUBA (B Manning)
PIP (Bellcore)
SIP (C Huitema)

Several new top level domains

verify that you can send mail to vse.cz

AOB

more than 30 hops between France and Japan (IP TTL recommended value is 64)

15.7. Routing Working Group (Tony Bates)

Tony Bates stood in for Jean-Michel Jouanigot who was unable to attend due to illness.

Closure on ripe-81

Gilles had a problem with blocks in the database - regarding who is actually responsible for splitting the block when only some of the networks are to be routed.

He proposes that the NCC will split the block accordingly to the requested routing policy. He thinks that it takes too much time/resource to ask a national registry to split the block.

Action: 15.18 NCC
To come up with a suitable solution to the problem of splitting blocks in the RIPE database when only some of the networks are to be routed

Antonio Blasco Bonito asked the question "what is the proper usage of the blocks?" It is currently not quite clear what the blocks are for. It may need some education for network managers. i.e. A block should be treated as a block.

Gilles Farrache commented that it would be impossible to tell all the network managers in Europe. Will be moved to CIDR point on the agenda.

The Chair moved to accept "ripe-81" as the official RIPE recommended method for registering routing policies within the RIPE database with the understanding that the update procedure would be modified to deal with blocks. THIS WAS AGREED and would be announced at the Plenary session the next day.

Ripe-81 enhancements

Tony Bates gave a presentation of some ideas for extensions for ripe-81.

Inter-AS Local Information

A way to represent local Inter-As information is needed (2 lines between 2 ASes)
Multi-homed multi inter AS connection

Seen as a need but outside of RIPE-81 itself. Peter Lothberg was volunteered to draft something.

Action: 15.19 Peter Lothberg To draft something on Inter-AS local information.

DMZs (DeMilitarised Zone)

DMZs are always an interesting case as generally the network is announced by more than one AS the issues are:

What AS should an DMZ be in?
Should the network be announced? not always? If it needs to be announced, you rely on others to announce it
should it represented in line with the routing protocol?
should it be in it's own AS? For Route Servers yes. "ptrace" issue
possible solution is a DMZ attribute in the RIPE database.
possible idea is to put in the database that this network is a DMZ with AS number and the list of AS'es that meet.

Action: 15.20 Tony Bates and Daniel Karrenberg
To draft possible attributes for DMZ Networks.

IP "Next Generation"

The idea of incorporating other protocols based on "ripe-81" formats. The topic was moved to next meeting.

Colouring

Havard presented his problem with EUnet/NO. The answer is to split the AS, but this cannot be done because of one router if you could source the nets in different ASes inside one router you could create different ASes with different policies. Other solution, take two links with different ASes

Action: 15.21 Daniel Karrenberg.
To draft document on colouring.

CIDR

Tony Bates presented one slide which stated:

"CIDR, default or die !!!!!!!"

He expressed his grave concern that network operators need to be ready and must understand the implications. It was clear that a document is needed detailing the impact of CIDR depending of what type of routing technique and topology you use. However, it was felt that the NCC could not take this action on themselves at present due to too many other documents in the pipeline.

EBONE will start first with CIDR so they have the responsibilities to document the consequences.

Stephan Biesbroeck (BelNET), Duncan Rogerson (JANET) and Harvard Eidnes (UNINETT) volunteered to detail how they see CIDR deployment will directly affect them. It is hoped this can be a basis for some form of CIDR impact document within Europe.

Action: 15.22 Stephan Biesbroeck, Duncan Rogerson and Harvard Eidnes
To evaluate in detail how CIDR deployment will directly affect them.

IXF BoF

The last 30 minutes was devoted to a discussion session on the ideas of the Internet eXchange Facility (IXF) as presented by Peter Lothberg.

Some limited copies of a draft by Tony Bates and Peter Lothberg entitled "IXF - An Interconnect Model for Internet Service Providers and Network Operators". The general feeling was that this was a good idea and should be explored more. The authors would modify the document in terms of context so it was clear what the aim was. Is was hoped that there would be continued discussion on the IXF issues at future meetings.

15.8. EBONE Report (Bernhard Stockman)

15.8.1. General update (Bernhard Stockman)

An overview was given of the recent developments in EBONE.

15.8.2. EBONE in East and Central Europe (Wilfried Woeber)

A very short overview of the current situation of EBONE connectivity for central Europe was given. For the time being Austria is "kind of hiding" these regions behind ACOnet. Therefore a project was started to propose a remodelling of this region to have an EBS in Austria and to connect the networks in this region as individual RBSs.

A proposal was submitted to the EBONE Management Committee to consider the restructuring of the dual-homed ACOnet RBS to become a true EBS. RBS connections are expected from Poland, Czechia, Slovakia, Austria, Hungary,Croatia and Romania (in geographical order).

The project milestones are:

April 7, 1993: on a request from the EMC, the EBONE Action Team reviewed and approved the proposal from the technical point of view
May 4, 1993: a costed proposal was submitted to the EMC for approval and for recommendation to the EBONE Committee of Contributing Organisations
June 29, 93: approval is expected on the ECCO level for 1993
Sept./Oct. 93: restructuring could be fully implemented, given ECCO approval is given in June
"in parallel": some reconfiguration activities for lines could already be started (where no formal decision is necessary) In the meantime considerable progress has been achieved!

15.8.3. EBONE 93/94 routing plan (Peter Lothberg)

An overview of the EBONE 93/94 routing plan was given by Peter Lothberg.

15.8.4. EBONE Traffic report (Willem van de Scheun)

The EBONE statistics report concerned the following topics: a review since the last RIPE meeting, IP accounting and traffic measurements.

IP Accounting

All EBS's now have CSC
Test gathering IP accounting has been done on January 20 and 21 and on March 23 and 24
Since April 1 all EBS's do IP accounting
To present some recent results data for Patrol 21 and 22 have been gathered and analysed

The results of the statistics for RIPE/non-RIPE aggregation taken on April 21 and April 22 were presented. Statistics for country aggregation are also collected and were presented for the same days.

Traffic measurements

Currently work is done at the OPSTAT format at Cern but there is a project by FARnet and Merit to develop traffic measurement tools.

Since April 15 for all EBS-RBS and EBS-EBS connections ifInOctets and infOutOctets are gathered
Aggregation is done on a daily basis
Collection of data form each EBS is done a weekly basis
Uses yet another, "simple" intermediate, format

It is possible to determine for example: How much traffic is flowing on EBONE; how much traffic is each regional delivering to EBONE and how much traffic is each regional getting from EBONE. Some data was also presented for the Amsterdam-EBS taken on April 27.

Please contact Willem van de Scheun or the RIPE NCC if you are interested in a copy of this presentation.

16. Next RIPE meetings

The 16th RIPE meeting was agreed to take place in Amsterdam on September 15th-17th, 1993. The RIPE chairs and the NCC agreed to investigate possible future venues for forthcoming RIPE meetings and to report at the 16th RIPE meeting.

Action: 15.23 RIPE chairs and NCC
Investigate possible future venues for forthcoming RIPE meetings and to report at the 16th RIPE meeting.

17. AOB

There were no items raised under this AOB.

18. Closing

Rob Blokzijl thanked the participants for attending and declared the15th RIPE meeting closed.

Appendix

List of Open Action Items

Action: 13.3 Marten Terpstra
Changes in database format. Coordination discussions with US counterparts are ongoing.

Action: 13.4 Marten Terpstra
Progress GSI/Merit/RIPE NCC exchange format discussions.

Action: 13.6 NCC
Include the operational contact attribute in all object definitions besides persons.

Action: 14.4 NCC
Generation of Internet handles by local registries.

Action: 14.17 Milan Sterba
Track the US development of the resource-object description work and will report to the list (see also action 15.9)

Action: 15.1 NCC
Implement update procedure with respect to ripe-81

Action: 15.2 Marten Terpstra
ripe-draft-in-addr-procedures-v1.4.txt to be archived as ripe document

Action: 15.3 Ruediger Volk
To suggest text on the necessity for organisations to supply details of their planned address space usage when requesting additional address space.

Action: 15.4 Bob Day
To fold in the comments of the working group on the ripe-draft-hints-v1.txt, circulate the document to the wg list and publish as a ripe document

Action: 15.5 Anne Lord
To fold in comments to the ripe-draft-hints-v1.txt and publish the leaflet "Delegated Registry"

Action: 15.6 RIPE chairs,chairs of working groups and RIPE NCC manager
Meet to progress RIPE NCC funding work item

Action: 15.7 Marten Terpstra
Fold in comments and republish amended ripe-72

Action: 15.8 NCC
Collect interesting cases of IP network number allocations (all local IR's, NCC) in preparation for tlhe local-ir closed seminar

Action: 15.9 NCC
To assist SURFnet for the testing and support of MBONE for Amsterdam IETF

Action: 15.10 Daniel Karrenberg
To propose new tags "created" and "assigned" to the databse working group for consideration.

Action: 15.11 Wilfried Woeber
Initiate discussion on the db-wg list proposing appropriate structure of the documentation for the database

Action: 15.12 Milan Sterba
to research and refine proposal for the "Resource Object" tag (see also action 14.17)

Action: 15.13 NCC
CEEC Report version 6 will be formatted into index data store with a hypertext tool installed which can be used to search the report.

Action: 15.14 Milan Sterba, Rob Blokzijl, Tony Bates
To coordinate and review the submissions for version 7 of the Connectivity Report.

Action: 15.15 Milan Sterba, NCC, RARE SEC
To produce version 6 of the report. NCC to print it, RARE SEC to copy and distributed by RARE SEC/NCC

Action: 15.16 Tony Bates
Publish straw 2 paper for comments and request volunteers to undertake some of the work.

Action: 15.17 Glenn Kowack and Tony Bates
Write a top level introduction for the user as an add-on to the GISS paper once it has been written.

Action: 15.18 NCC
To come up with a suitable solution to the problem of splitting blocks in the RIPE database when only some of the networks are to be routed.

Action: 15.19 Peter Lothberg
To draft document on Inter-AS local information.

Action: 15.20 Tony Bates and Daniel Karrenberg
To draft possible attributes for DMZ Networks.

Action: 15.21 Daniel Karrenberg
To draft document on colouring.

Action: 15.22 Stephan Biesbroeck, Duncan Rogerson and Harvard Eidnes
To evaluate in detail how CIDR deployment directly affects each of them.

Action: 15.23 RIPE chairs and NCC
Investigate possible future venues for forthcoming RIPE meetings and to report at the 16th RIPE meeting.