Rob Blokzijl welcomed the participants to the 11:th RIPE meeting.
Bob Day (JANET/JNT, R.Day _at_ jnt.ac _dot_ uk) sent his apologies.
Agenda for the 11:th RIPE meeting. Minutes from the 10:th RIPE meeting. RIPE DNS Hostcount. RIPE DNS Hostcount History. RIPE Database Status Report. A new RIPE Working Group on the Relations Between the Research and Academic Networks and the Commercial Networks. Overview of East and Central European Networking. Policy Based Routing within RIPE. A RIPE recommendation on Routing Policies. RIPE and COSINE MHS service coordination topics. Draft EBONE Action Team Implementation Plan.
The proposed agenda was approved. Logistics and the organization for this meeting was decided.
2. Minutes of the last meeting.
The minutes from the 10:th RIPE meeting were approved.
Review of the action list:
- Produce a document on raw data formats for statistical data. ACTION CLOSED. The IETF OPSTAT work will be recommended for use.
- Produce a document on possible NNstat sites among RIPE connected networks. ACTION OPEN. Will be discussed in the Monitoring and Statistics Working Group.
Tony Bates and Marten Terpstra -
- Produce a document on experiences of running IP/IXI. ACTION OPEN.
- Contact CISCO to establish formal contacts. ACTION CLOSED. CISCO was contacted at the Santa Fe IETF.
- Review documents to be installed at the nic.eu.net document archive. ACTION CLOSED. The archive has been updated. - Investigate possible restructuring of the RIPE WGs and areas of interests. ACTION CLOSED. A proposed action will be described on Wednesday. -
- Produce a document regarding the group to which the RIPE NCC should report back. The document should be distributed to the RIPE mailing list for discussions. ACTION CLOSED. The subject is closely related to the previous action.
- Produce a recommendations document on security issues. ACTION OPEN. - Coordinate preparations for a NMS demonstration at the next RIPE meeting. ACTION CLOSED. - Prepare a NetBlazer demonstration at the next RIPE meeting. ACTION CLOSED.
- Produce a document on statistical metrics to gather and how to store this data. ACTION CLOSED. The IETF OPSTAT work will be recommended for use.
Francis Dupont and Milan Sterba.
- Produce a cookbook on DNS. First version ready at the next RIPE meeting. ACTION OPEN.
- Organize the policy based routing implementation for RIPE connected networks. ACTION CLOSED.
- Write and make easily available a cookbook about European network information services. ACTION CLOSED. Replaced by new RIPE working group initiative.
- Check for RIPE connected areas not receiving CERT bulletins. ACTION CLOSED.
- Make a new distribution of the database tools available. ACTION OPEN. Work is progressing.
- Coordinate preparations for a Dial Up IP presentation at the next RIPE meeting. ACTION CLOSED.
- Distribute the NORDUnet NMS checklist to RIPE. ACTION CLOSED.
- Distribute the draft RIPE statistics document to the ripe-org mailing list for further comments. ACTION OPEN. The IETF OPSTAT document will be distributed when ready.
- Send a list of proposed but not installed documents to Rob Blokzijl. ACTION CLOSED. The archive is already up to date.
- Write a RIPE section to the Internet Monthly Report. ACTION TRANSFERRED. Rob Blokzijl and Daniel Karrenberg
- Inform Matti Rendahl on needs for DNS statistics. ACTION CLOSED. DNS Statistics is available. -
- Distribute the IETF OPSTAT papers to the ripe stat mailing list. ACTION OPEN. Awaiting the IETF OPSTAT to complete its work.
- Produce a report from the Atlanta IETF. ACTION OPEN.
- Produce a document on methods for incremental updates of the RIPE database. ACTION OPEN. - Produce a report from the Atlanta IETF. ACTION CLOSED.
3. RIPE Technical Session Report.
A technical session was for the first time arranged on Thursday and Friday the week before this RIPE meeting. Three major areas were covered: low cost IP technology, IP over ISDN and SNMP based management stations. This combination had some drawbacks. The technical "latest and greatest" dominated too much over the tutorials.
Connectivity was demonstrated with IP/ISDN, IP/X.25 and IP/Serial line. Bugs were found as well as features that were needed. The ISDN connectivity included an ISDN switch from Philips.
Four SNMP management stations were demonstrated.
The conclusion is that the technical session should be repeated but there should be a clear differentiation between tutorials and other activities. A report will be written and the participants are asked to send in comments to Daniel Karrenberg for inclusion in the report.
4. Short reports.
The COSINE MHS project suggested to have an informal meeting between them and RIPE, in order to explore area's of common interest. Such meetings will be held the second day of this RIPE meeting.
5. Connectivity issues.
5.1 Western Europe.
Report from Mike Nowlan, EUnet.
Beginning in December, 1991, the SWITCH network in Switzerland closed its network to traffic to EUnet users.
This action was taken without prior notification, resulting in significant wasted technical and managerial effort. As far as we can determine, SWITCH has never published a policy on this issue.
We request that RIPE make a general recommendation regarding advance notification by network operators of any action which might impair network connectivity.
We also ask RIPE to encourage participating organizations to publish their policies for connection to other networks. Without such clear policies the future progress of European networking may be impeded.
The EUnet request is partially fulfilled by the RIPE whois database, including information on policy. The problem should be taken up within RARE as a policy question, not in RIPE which is a technical body. However, RIPE strongly recommends that changes of policy should be announced in due time.
5.2 Central and Eastern Europe.
Report from Milan Sterba, INRIA, on East European Issues.
Besides the paper on ECE networking Milan reported on a draft proposal from Klaus Ullman, DFN, to the CEC for funding of an extension of the current IXI network into the East European countries. The requested funding was for 6.8 MECU.
It is seen as a well appreciated goal to help the East European countries with connectivity but this effort confuses the use of technology.
6. Relations between Academic and Research networks and Commercial Networks
Introduction of a new RIPE activity.
Glenn Kowack, EUnet, described a proposed new RIPE activity regarding the relations between research and academic networks and commercial networks. A paper describing this activity has been distributed.
7. Network Information Services.
7.1 Report from the IETF User Services Working Group.
Joyce Reynolds, ISI, reported from the IETF User Services Working Group.
User Services Planning in the Internet
- ISOC = Internet SocietyIAB= Internet Activities Board
- IRTF = Internet Research Task Force
- IETF = Internet Engineering Task Force
- RFC = "Request for Comments" document FYI = "For Your Information" document
7.1.2 Internet Society
- Organizers: CNRI, EDUCOM, RARE
- Seeking Other Interested Parties
- General Operations: Started 1 Jan 92 * Individual and...
- Founding Organizational Members are invited to join now
* Multiprotocol Evolution. TCP/IP, OSI,... * Encourage Internet Growth * Educate the Public * Stimulate Provision of Service * Recognize Individual Contribution * Promote and Explore Scientific, Educational, Business Use * Facilitate Collaboration
* International Professional Society * Individual and Organization Members * Elected Board of Trustees who appoint officers * Staggered 3-year terms, max 2 contiguous terms * Incorporates IAB / IETF / IRTF
* Evolution of Internet Technology * Incorporate IAB, IETF, IRTF * Newsletter and Journal * Annual Conference: INET, INET 92: June 15-19, 1992 Kobe, Japan * Possible Infrastructure Assistance - CERT-System - Crypto-Certificates - Internet Registry - Referral Services
* Nourish the Community now growing up around networks. * Our packets cross borders freely and our sense of community should be equally open. * Continue the Grand Collaboration now linking 5,000+ nets and 3,000,000+ people.
FOR MORE INFORMATION...
ISOC _at_ NRI.RESTON.VA _dot_ US FAX: +1 703 620 0913 TEL: +1 703 620 8990
7.1.3 IETF User Services Working Groups and Projects
Directory Information Services Infrastructure (DISI) Chaired by Christopher Weider
DISI chartered to facilitate deployment of X.500 Directory Services on the Internet, by producing "Administrator's Guides".
Current working Internet-draft: "Executive Introduction to Directory Services using the X.500 Protocol".
Internet Anonymous FTP Archives (IAFA) Chaired by Peter Deutsch and Alan Emtage
IAFA is chartered to define a set of recommended standard procedures for the access and administration of anonymous ftp archive sites on the Internet.
Internet School Networking (ISN) Chaired by John Clement, Art St.George and Connie Stout
ISN is chartered to facilitate the connection of the United States' K-12 (Kindergarten-12th Grade) schools, public and private, to the Internet, and promote school networking in general.
Network Information Services Infrastructure (NISI) Chaired by April Marine and Patricia Smith
NISI is exploring the requirements for common, shared Internet-wide network information services. The goal is to develop an understanding for what is required to implement an information services "infrastructure" for the Internet.
Current working Internet-draft: "Building a Network Information Services Infrastructure"
NOCTools Chaired by Robert Enger and Darren Kinley
"Son of NOCTools" is updating and revising their catalog to assist network managers in the selection and acquisition of diagnostic and analytic tools for TCP/IP Internets.
User Documentation (UserDoc) Chaired by Lenore Jackson and Ellen Hoffman
User-Doc is preparing a revised bibliography of on-line and hard copy documents/reference materials/training tools addressing general networking information and how to use the Internet. (Target audience: those individuals who provide services to end users and end users themselves.)
User Glossary (UserGloss) Chaired by Gary Malkin and Tracy LaQuey Parker
User-Gloss is chartered to create an Internet glossary of networking terms and acronyms for the Internet community.
User Services (USWG) Chaired by Joyce K. Reynolds
The User Services Working Group provides a regular forum for people interested in all user services to identify and initiate projects designed to improve the quality of information available to end-users of the Internet.
User Services Area Council (USAC)
USAC is responsible for researching and defining short term and long term user services needs internationally, and coordinating developments in finding solutions.
Philosophy for Membership
1) People are selected as individuals, NOT representatives of organizations, countries, networks, or anything else.
2) We like to have people with a variety of backgrounds and experience.
IETF User Services Area Plan
Seven types of User Services objectives:
- User Information
- Network Information Services Infrastructure
- Network Operational Management
- Education - Documentation and Distribution - Interaction with IESG Areas
- Interaction with other international user services entities
- User Information
The Internet community requires up-to-date, basic Internet knowledge and experience. These can be achieved by publication of handbooks, bibliographies, directories, and glossaries. Yet, how does the IETF "get the word out" beyond the normal distribution and announcement via the RFC series??
Identification and research on various existing distribution resources, and consideration of possible long term distribution methods are required.
Network Information Services Infrastructure
A global infrastructure for common shared Internet-wide network information services is needed. Research and development for an information services "infrastructure" for the Internet community. Documentation of methods for the interaction and cooperation among NICs. We intend to coordinate closely with other efforts in the international networking community.
Network Operational Management
This topic overlaps with other IETF Areas such as network management, operations, and applications. Yet, development of general information to users is essential. The User Services Area intends to contribute by providing documentation that will be developed in tandem with technical specifications.
The educating of new network users is mandatory. For teachers, trainers, etc., provide organized Internet tutorials, "hands on" training programs, active participation in the K-12 education initiative, and Internet specific documentation.
Documentation and Distribution
Coordinate the development of informational documentation and distribution methods for the Internet community. FYI RFCs are introductory and overview documents for network users. Their purpose is to make available general information, rather than the protocol specifications or standards that is typical of other RFCs.
Interaction with IETF Areas
Coordination with other IETF Areas to work together on topics of common interest.
Interaction with other international user services organizations.
- Helping New Users
- Task for the Networks (e.g., Regionals)
- New User Guides -Etc., etc.
IETF User Services is second level. Dealing with real users is first level. IETF User Services provides information to people doing first level services.
Promote information sharing and cooperation between Network Information Centers.
SRI NISC Merit BBN NNSC CERF-NIC Others
Internet Documentation for USERS (not WIZARDS!):
7.1.4 FYIs and RFCs
FYI 11 "A Catalog of Available X.500 Implementations", Jan. 1992.
FYI 10 "There's Gold in them thar Networks! or Searching for Treasure in all the Wrong Places", Dec. 1991.
FYI 9 "Who's Who in the Internet: Biographies of IAB, IESG and IRSG Members", Aug. 1991.
FYI 8 "Site Security Handbook", Jul. 1991.
FYI 7 "FYI on Questions and Answers: Answers to Commonly Asked "Experienced Internet User" Questions", Feb. 1991.
FYI 6 "FYI on the X Window System", Jan. 1991.
FYI 5 "Choosing a Name for Your Computer", Aug. 1990.
FYI 4 "FYI on Questions and Answers: Answers to Commonly asked "New Internet User" Questions", Feb. 1991."
FYI 3 "FYI on Where to Start: A Bibliography Internetworking Information", Aug. 1990.
FYI 2 "FYI on a Network Management Tool Catalog: Tools for Monitoring and Debugging TCP/IP Internets and Interconnected Devices", Apr. 1990.
FYI 1 "F.Y.I. on F.Y.I.: Introduction the F.Y.I. Notes", March 1990.
- Started in 1969 - Now have 1292 - Since RFC 500, all are available on-line - Distribution is primarily via the network (FTP & EMail)
Internet Documentation Process
1) Final documents are RFCs. Available on-line in libraries around the world.
2) IETF Working groups develop drafts. IESG reviews the drafts, makes recommendations. IAB approves the drafts for RFC publication.
7.1.5 Future Plans
- Common Questions & Answers in the Internet - Common User Interfaces and User Perspectives - Copyright & Intellectual Property - Distribution & Announcement of IETF products - Historical Documentation - Interface with Commercial Networks - Network Service Guide - "Public Relations" for the Internet
Network Information Services Infrastructure
- Interaction & Cooperation among NICs - Internet Administrator's Guide - New Applications to make NICs Work More Efficiently. - User Liaison Guide
Network Operational Management
- Ethics and Etiquette - Internet Installation Checklist - Internet Introduction Packages - Site Policy Handbooks - DNS "Cookbook"
- Education of Special Groups - IETF Tutorials - K-12 Education Initiative - "Standard Curriculum" Training Programs and Improving Training Material for Local Instructors - TCP/IP Tutorials
More cooperation between RIPE & IETF
7.2 Start of a new RIPE activity.
Nandor Horvath described a new RIPE activity in the area of network information services.
8. Network Management.
8.1 Report on the status of the RIPE NCC.
Rob Blokzijl reported on the status of the RIPE NCC. The last RIPE meeting was just before the RARE CoA meeting were it was decided to implement the RIPE NCC. Since then an advertisement on the solicitation for the RIPE NCC manager and site has been written and distributed to the RIPE and RARE mailing lists. The period for bidding ended in October. The original plan for the RIPE NCC selection panel had been to include the RIPE chairman and Phil Jones (RIPE member) at the four member panel, to be chaired by Professor van Binst and with Lars Backstrom as a fourth member. As NIKHEF had applied, Rob Blokzijl has been replaced on the panel by Yves Devillers.
Yves Devillers reported from the work in the RIPE NCC selection panel. The final recommendation for the RIPE NCC manager and site has been unanimous decided by the selection panel. The content of the recommendation is still confidential until approved by the RARE EXEC which will meet on January 29.
The RIPE NCC will be installed and functional at earliest in March.
[ Editors Note: A few days after this RIPE meeting it was announced that the RARE CoA accepted the recommendation of NIKHEF in Amsterdam as the RIPE NCC site and Daniel Karrenberg as RIPE NCC manager.]
8.2 Report from the Internet NIC.
Scott Williamson, GSI, reported from the Internet NIC. The two major problem recently in connecting to the NIC were the connectivity bottleneck due to the low bandwidth line within MILNET and the Ingres database implementation. Both problems have now been resolved.
Clearing up of some confusing points in the application template will be done. The IP registry problem with block allocation has to be resolved. The NIC will stay out of politics which should be worked at by involved countries.
The NIC is managed by Network Solutions which is subcontracted by GSI for this. Another major area for Network Solutions is to act as security coordination center in Milnet.
The administration of IP addresses and AS numbers was delegated by Jon Postel to the NIC. There is an RFC describing this. Other IANA responsibilities such as protocol and well known port numbers are still maintained within IANA.
RIPE and the NIC will together discuss the formation of an European IP registry.
8.3 Report from Merit.
Elise Gerich, Merit, reported from the current status in the NSFnet.
9. DNS issues.
Francis Dupont, INRIA, expressed the need for an updated version of the ISO 3166 standard for two letter country codes. Hank Nussbacher volunteered to get a copy but Larry Landweber probably has a copy online.
The problem with very small domains was discussed. The general approach should be to allow small domains. The problems which may arise will be discussed within the DNS working group.
10. Routing Issues.
Jean-Michel Jouanigot presented current issues for the RIPE routing working group. There will one closed and one open routing working group session. A new version of the RIPE document on the need for policy based routing is distributed. There is a need for two new objects in the RIPE whois database. It is necessary to define a language for expressing policies and to construct tools for converting these polices into router configuration files.
Areas of interest to be covered:
- Find solution for routing policies for each center. - Current BGP policy based routing test shall be reviewed. - IP/IXI routing issues.
The RIPE document entitled "A RIPE Recommendation on Routing Policies", distributed during this meeting, should be renamed to "A RIPE Recommendation on Connectivity Policies". This should not be interpreted that RIPE sanctions whatever kind of policy and especially not the decreasing of current level of interconnectivity between RIPE members.
11. Network Monitoring and Statistics.
Bernhard Stockman, NORDUnet, presented actual topics for the Network Monitoring and Statistics working group. The IETF OPSTAT work on a model for Operational Statistics is planned to soon be transferred into a Internet-Draft. A model for Trouble Ticket exchange formats will be discussed based on the current IETF UCP working group Internet-Draft. There is an urgent need to produce detailed traffic analysis in crucial interconnection points between RIPE connected networks. The NNStat deployment within the RIPE community will be treated as a possible way to achieve such statistics.
The Monet and Monster network management tools will be presented.
12. Presentation of the current EBONE-92 status.
Bernhard Stockman presented the latest news from the EBONE-92 initiative. The EBONE Exec met in Amsterdam on January 21, where the proposed EBONE-92 implementation plan where discussed. The proposed 2 Mbps X.25 line between UK and Germany will be replaced by a line between London and Montpellier. The reason for this is the still unclear situation for deployment of this line. The EBONE-92 is an urgent effort and shall be implemented without unnecessary delays. Germany is free to connect to the EBONE as a RBS for the time being and as EBS as soon as confirmed international connectivity is in place.
13. European Status Information.
Marten Terpstra, NIKHEF, reported on the current DNS hostcount and presented a history on the development of the DNS hostcounts. See distributed papers.
13.2 RIPE whois statistics.
Daniel Karrenberg presented the current status of the RIPE database. See distributed paper.
Of the current 1400 non-US networks in the NSF database around 900 is European which is 25% of the complete NSF database with around 4000 networks. Of the 900 European networks 20% is class B and 80% class C networks.
14. Working Group Session Reports.
14.1 Eastern and Central Europe. (Milan Sterba).
Report from the European Connectivity WG
The work of the European connectivity WG was completely dedicated to Eastern and Central European connectivity. The issues relevant to Western Europe having been tackled within the Ebone 92 framework.
Substantial updates to the ECE connectivity report draft have been made. The final version of the document will be published on the RIPE list and placed in the RIPE documents repository on nic.eu.net.
Many international network supporting activities are currently underway, RIPE is acting as an information interchange and informal coordinator of a number of them
A proposal of extending the existing IXI infrastructure to ECE countries in the framework of an EEC project has been considered with a lot of interest and has been extensively discussed. It has been stated that a coordination with R&D networking initiatives in the ECE countries and their current plans in national and international networking was a necessity. The expertize of the west-European networking specialists materialized in the EEPG report and the RARE Operational Unit project as well as the most recent Ebone 92 evolutions should be taken into account in every such project in order to avoid repeating of known errors and to consistently integrate ECE countries into the actual European networking context. The war of protocols should be avoided at any price.
On Feb 13-14th on the occasion of the official opening of Internet services in Czechoslovakia, ECE countries representatives agreed to meet in Prague in order to coordinate their R&D network strategies and discuss issues related to common network infrastructure, training programs, network services and operation.
First proposals of ECE countries cooperation have been discussed (CS, PL, HU backbone, network summer school in one of the ECE countries etc.)
A strong interest exists in the ECE countries to put an Ebone Boundary System to Austria, due to the important role Austria is playing in the international connectivity.
A solution to the persisting licensing problems with CISCO routers will will be sought for by contacting persons responsible for licensing at CISCO US by Milan Sterba
An action dedicated to sensiblizing ECE countries networkers to ISOC membership will be lead by RIPE in coordination with ISOC and using RIPE diffusion channels
A smaller group met after the session of the WG to make a list and an evaluation of existing (and missing) cheap IP solutions, which could help spreading IP connectivity over ECE countries. The output of the group will be distributed in a separate document.
14.2 Relations between Academic and Research Networks and Commercial Networks. (Glen Kowack).
[This summary statement was made by group chair Glenn Kowack, to the plenary session of the RIPE meeting of Wednesday, 22 January 1992.]
The group plans to continue to provide a forum for discussion between these two facets of European networking.
We drafted a preliminary charter, a list of goals, and a list of problems which require our consideration.
To produce recommendations for RIPE as necessary to enable smooth interworking between commercial networks on the one hand, and research, academic, and education networks on the other.
- worldwide - fully-interconnected - available
- Appropriate controls to support growth - technical progress - facilitation of organizational evolution - (possibly) provide arena for arbitration
- Preliminary Problem List fragmentation
- imbalance of representation - unfair competition - routing - no vehicles for cost sharing - no agencies for central service provision - no standards for acceptable use policies
We discussed our eventually contacting other organizations with similar charters in other regions, including the IEPG, and the IETF's ORAN (Operational Requirements Area Directorate).
We consider it vital that all major IP network operators be involved in this dialog, especially the commercial networks. We will make efforts to involve these groups.
14.3 Network Information Services (Nandor Horvath).
In this working group session two kinds of participants were present; those who needed advices and other who gave advices which the first group was greatful for. It was clarified that the information services are needed at two different levels: the level for network experts and the level for end users. Not to reinvent the wheel the gained experience with existing services in the US, in NORDUnet, in France, in Australia and in the UK (Newcastle) could be studied first as reference. We learned also that the RARE WG3 is dealing with similar problems and it will be useful to contact them.
After studying what these groups have done in their fields we will decide what we can use from their work and if there is something new we should add to this.
14.4 Network Management (Rob Blokzijl).
RIPE, NIC and Merit described their current databases, and procedures for updating them. From here followed the following areas for discussion:
- harmonization of the databases: they should contain the same information on the same objects. This is not always the case today. Therefore a 'cleaning up' operation has to start as soon as the RIPE NCC comes into operation.
- exchange format of database entries. In order to keep the databases in sync, regular exchange of updates will take place. For this we need an exchange format. Again, an action to be taken up by the NCC.
- it was also decided that RIPE and the NIC would exchange full dumps of their databases at regular intervals.
- - concern was expressed on the security aspects of database exchanges. Concluded was that the progress on security enhanced mail systems should be closely followed. In the meantime the database maintainers are strongly adviced to continue archiving all documents pertaining to database change requests.
- - it was strongly recommended to construct the RIPE databases in such a way that they can be easily transferred into X.500 directory systems. The COSINE Paradise project will be contacted to exchange ideas.
14.5 DNS Issues (Francis Dupont).
14.5.1 Very Small Domains.
National domain managers are faced to the problem of domains for very small entities and should solve two problems :
- how to split the name space to allow for technical and administrative manageability
- how to delegate parts of the name space
- Two general rules have been stated again :
- domains are not limited to organization (i.e. a private person can ask for a domain)
- customers of a service provider must not be housed under the domain of the service provider (this domain is for infrastructure (NIC, NOC, routers,...) not for clients).
A geographical scheme can be used for small schools (K-12 in USA), for instance the schule.de domain in Germany, following the educational organization.
The problem is much harder when it goes for private persons since geographical schemes do not allow for easy delegations of wide subsets of the name space.
It has been suggested that private person domains should be housed under a dedicated sub-domain of one country (e.g. pp.at with mozart.pp.at).
14.5.2 Multinational Domains.
Two solutions have been proposed for registering domains of multinational entities:
- one domain per country but it is much work and it is not possible to guarantee the same name in all the countries (possible clash with already existing names, different naming policies, ...)
- a multinational domain (net, com, org, int : some documents describing the usage of these multinational domains should be available).
14.5.3 Presentation of COSINE-MHS
Claudio Allochio reported on experiments with a DNS implementation of RFC987 tables in Italy.
A strong coordination between DNS national and X.400/X.500 naming authorities is recommended. For example when a second level DNS domain is registered the RFC 987/1148/1148bis mapping for this domain must be provided.
Two important contacts (E-mail addresses):
technical questions about RFC 987:
where to get mapping tables:
(with FTP / FTAM on nic.switch.ch)
14.6 Cosine MHS.
A request has been received from the COSINE MHS project to investigate areas of common interest between the project and RIPE technical groups. Two area's were identified:
- The use of the DNS name space by the MHS project. This will be a subject for study by the DNS specialists. (See agenda point 14.5.3 above).
- The use of the RIPE 'whois' database for storage of operational data for the MHS project (description of contact persons, description of MHS gateway systems, etc.) This will be taken up by the database group.
14.7 Routing Issues (Jean-Michel Jouanigot).
14.7.1 Policy based routing
The document 'Policy based routing within RIPE' was discussed during the closed session, and the definition of two new objects in the RIPE database was finalized. These documents were presented and commented during the opened session. The 'rout-pr' field definition was changed, and it was agreed to add a description of the procedure to get a privilege, as well as an example of Policy based Routing implementation. The final version of these documents should be published soon.
Tools to use the new database fields are needed. A "Policy Compiler" is currently studied at CERN. The first version of this tool should be available in two month time.
The Database Management is kindly asked to implement the new database objects, as soon as the final version of the definitions are agreed.
The Routing Policies of the routing centers are to be collected. It was proposed to collect them via e-mail as soon as the document is finalized. For each routing center, the following information will be published:
- Which networks are accepted on each router interface
- Which networks are redistributed on each router interface
- Which 'default routing' policy is used, and how the routing center routes to the U.S.
There is now an urgent need for a BGP pilot. Information from the IETF and some GMD tests confirm that:
- Cisco 8.2(6) is the best release for BGP-I, but does not implement Autonomous System Path filtering.
- Autonomous System Path filtering (as well as BGP-II & III) are implemented in release 9.0 of the Cisco software. The version is currently under Beta test.
Several organizations are willing to participate in a BGP pilot in Europe, in collaboration with our U.S. collegs if possible:
Cristina Vistoli GARR INFN vistoli _at_ cnaf.infn _dot_ it Dave Morton ECRC Muenchen dave _at_ ecrc _dot_ de Francis Dupont INRIA Paris Francis.Dupont _at_ inria _dot_ fr Gilles Farrache IN2P3 Lyon Farrache _at_ frcpn11.in2p3 _dot_ fr Jean-Michel Jouanigot CERN jimi _at_ dxcoms.cern _dot_ ch Juha Heinanen FUNET Juha.Heinanen _at_ funet _dot_ fi Micheal Ernst HEPnet DESY R02ERN@DHHDESY3.bitnet Miguel A. Sanz RedIRIS Madrid miguel.a.sanz _at_ iris-dcp _dot_ es Stefan Fassbender EASInet GMD stf _at_ easi _dot_ net Tony Bates JANET ULCC T.Bates _at_ noc.ulcc.ac _dot_ uk
A mailing list, has been created.
The goals of the pilot are to be defined as well as the topology of the routers/links involved.
14.7.3 IP over IXI
Marten presented the situation of the NIKHEF IXI access point. The 64k line is overloaded for quite some time now, and the quality of service is rather bad. Solutions have to be found to improve the situation.
14.8 Network Monitoring and Statistics (Bernhard Stockman).
A review of the current status in the IETF OPSTAT WG was done. The work is almost finished and a Internet-Draft will be distributed within shortly. At the last IETF OPSTAT meeting in Santa Fe two major areas were treated; how to store statistical data and which polling period to use. Regarding the polling period it has been suggested to use a scheme with varying polling periods for different metrics. Some metrics may need tight polling to cover peak behaviors. The shortest period suggested is 60 seconds. To avoid excessive needs for storage the polling scheme was accompanied with a scheme of preprocessing and aggregating of stored raw statistical data.
There is a need to produce detailed traffic analysis in crucial interconnection points between RIPE connected networks. For this reason NNStat is believed to be an adequate tool that should be investigated. Merit has announced that their NNStat log file processing tools now have been made public available. One example of a interconnection point of interest is the IBR-LAN at WCW in Amsterdam which carry traffic between a large amount of the RIPE interconnected networks. To be able to catch all the ethernet packets there is a need for at least a SUN Sparc Station. The RIPE NCC is believed to be the right body to when installed perform the actual installation and maintenance of the NNStat tools.
The working group discussed the need for a document on a RIPE statistical wish-list. RIPE connected network operations centers are encouraged to send in suggestions for such a list. Bernhard Stockman volunteered to act as editor of such suggestions to produce a wish-list paper.
The Working group decided not to treat the Trouble Ticket exchange formats due to time constraints.
James Barr presented the current status of the Monet Network management monitoring tool. A proposal for a demo at a comming RIPE meeting will be distributed on the RIPE mailing list. It was expressed a need for improvement of the current documentation which today is complex and hard to penetrate. It was recommended with some kind of easy to go tutorial just to get Monet up and running. There is a need for test sites. KTH, ULCC and WCW were suggested.
Daniel Karrenberg presented Monster (Monitor Station and Error Reporter). The monster tool consists of a central polling tool which based on the MIB database and a configuration file produces raw statistical data. From this raw statistical data different traffic analysis is possible to do using other tools. Examples of post-processing tools are a graphical plotting tool, an event extractor and a statistical data aggregating tool. The Monster tool is currently being used at SARA as event reporter. There is a need for more beta test sites. Please contact Daniel Karrenberg if you are interested. If any interested there is also a need for a trap daemon.
14.9 RIPE mapping working group.
There is currently no convener for the RIPE mapping effort. Hank Nussbacher volunteered to act as an interim convener until a new chair is found. Hank will write a charter for this working group to restart the work.
14.10 RIPE working group structure and relations to the RIPE NCC.
As have been discussed for some time there may be a need to investigate the current RIPE working group structure. The relation between RIPE and the RIPE NCC needs to be investigated. A group consisting of the RIPE chair, the RIPE working group chairs and the comming RIPE NCC manager will be formed to be responsible for both this issues.
14.11 Evaluation of the RIPE technical session.
The first RIPE technical session was held as a separate event during two days before the RIPE meeting. The combination of both connectivity demonstrations and tutorials were not a good thing as the first tended to dominate. There will probably be the need for a clear separation between these two activities in the future.
Erik-Jan Bos expressed the need for future network management demos and volunteered to put in more work to make this happen.
The next RIPE technical session was proposed for the fall RIPE meeting.
Daniel Karrenberg requested that if anyone notices an interesting vendor in the areas of Network Management, Routing and Statistics that this vendor should be asked to take contact with RIPE for participation in the Technical Session. Daniel volunteered to coordinate this efforts.
It was also expressed the usefulness of having the exhibition last until the morning at the first day of the RIPE meeting so that RIPE meeting participants arriving early may have a chance to visit the demonstration.
15. Reports from other groupings.
Rob Blokzijl reported on the status of the RARE Task Force on the Operational Unit. The RIPE NCC will not automatically be integrated in an Operational Unit. RARE being an organization for the R&D community have a narrower scope than RIPE. The organizational set up of the RIPE NCC has been agreed on by RIPE and can only be changed in collaboration with RIPE. The RIPE NCC will also play a role in the maintenance of the homogeneous Internet and by this be related to ISOC. Rob will write a paper on the history and scope of RIPE to make its position clear.
It was noted that the same parties who are now funding the NCC, are the potential funders for an Operational Unit and perhaps these parties would mind paying two entities who could possibly merge.
Bernhard Stockman reported from the last IEPG meeting in Santa Fe. The major topic of interest is currently the possibility of creating an robust Internet core in the same sense as being done in Europe with the EBONE-92 initiative. There has been proposed a model for interconnected Network Exchange points (NEX's) where network organization should be allowed to connect charged with a flat fee. The next IEPG meeting will concentrate on this efforts and is suggested to be the week before the IETF meeting in San Diego.
Rob Blokzijl reported from the last CCIRN meeting in Santa Fe. A paper for connectivity coordination was drafted by Barry Leiner. There has been a lot of commenting around this paper and a final version is not yet ready. The problem is the definition of a link, i.e. what is to be coordinated.
16. Date and place and time of next meeting.
The next meeting has been decided earlier to held in Amsterdam, April 27-29, 1992. The following meeting was decided to be at September 30 - October 2, 1992. The place was not decided but left open for proposals. It should be noted that any place being offered as meeting place must be able to also undertake the installation and running of the RIPE technical session.
17. Any Other Business.
17.1 Report on a new German Group.
Ruediger Volk reported on the formation of a new group in Germany, Deutsche Intressen Gemeinschaft Internet (DIGI). This group had its first meeting in Muenchen, December 6, 1991. A mailing list has been set up; . A ftp archive has been installed on the same host.
The objectives for DIGI are:
- Caring for the housing of a German NIC including the administration of the .de domain and the running of a delegated IP registry in close cooperation with the RIPE NCC.
- Helping users (on an administrative level). Get and exchange information both for general use and for IP technicians.
- Help to create and maintain a competitive and fair market for Internet services by keeping a proper full interconnectivity and keeping up of the traditional Internet spirit of cooperation.
- Put emphasis on the user side. - Organize overall user representation.
Both academic and commercial as well as other sectors will be addressed.
17.2 Proposal for RIPE meeting structuring.
It was requested that the RIPE meeting is structured in blocks of interest so that it should be possible to visit a RIPE meeting just for one day. The proposed RIPE group for investigation of the of the RIPE working group structure will look into this.
Rob Blokzijl thanked the all the participants and declared the 11:th RIPE meeting closed.
APPENDIX: Action List.
- Produce a document on possible NNstat sites among RIPE connected networks.
Tony Bates and Marten Terpstra
- Produce a document on experiences of running IP/IXI.
- Write a document on the history and scope of RIPE to make its position clear.
Rob Blokzijl and Daniel Karrenberg
- Write a RIPE section to the Internet Monthly Report.
- Produce a recommendations document on security issues. - Prepare for next technical session.
Francis Dupont and Milan Sterba.
- Produce a cookbook on DNS. First version ready at the next RIPE meeting.
- Make a new distribution of the database tools available. - Coordination of vendor presentations at the next RIPE technical session.
- Write a charter for the RIPE mapping working group. - Get a copy of the latest version of ISO-3166
- Distribute the draft RIPE statistics document to the ripe-org mailing list for further comments. - Distribute the IETF OPSTAT papers to the ripe stat mailing list. - Produce a report from the Atlanta IETF. - Coordinate and edit a RIPE statistics wish-list paper.
- Produce a document on methods for incremental updates of the RIPE database.