1. Opening Rob Blokzijl welcomed the participants to the 12:th RIPE meeting.
1.1 Papers tabled:
Agenda for the 12th RIPE meeting. Network Information Services DNS issues Routing issues Network monitoring and statistics gathering Europen connectivity RIPE database report RIPE position paper .DE
The proposed agenda was approved with the addition of a presentation to be given by DEC representative Ken Punshon on the technical details of ongoing developments of multiprotocol routers within Digital.
Logistics and organisational details for the meeting were finalised.
2. Minutes of the last meeting.
The minutes from the 11:th RIPE meeting were approved. Review of the action list: Tony Bates: - Produce a document on possible NNstat sites among RIPE connected networks. Action aborted.
Tony Bates and Marten Terpstra: - Produce a document on experiences of running IP/IXI. Action open.
Phil Jones expressed concern for international IP/IXI traffic. There is a need to investigate the technical configurations for an optimal service.
- Write a document on the history and scope of RIPE to make its position clear. Will be done during this meeting.
- Produce a recommendations document on security issues. Action open. - Prepare for next technical session. Action open.
Francis Dupont and Milan Sterba:
- Produce a cookbook on DNS. First version ready at the next RIPE meeting. Action open.
Milan asked for input on the content of the document.
- Make a new distribution of the database tools available. Action open. The work has been started.
- Coordination of vendor presentations at the next RIPE technical session. Action open.
Hank Nussbacher and Don Stikvoort: - Write a charter for the RIPE mapping working group. Action open
- Get a copy of the latest version of ISO-3166 Action completed.
- Distribute the draft RIPE statistics document to the ripe-org mailing list for further comments. Action open.
- Distribute the IETF OPSTAT papers to the ripe stat mailing list. Action completed.
- Produce a report from the Atlanta IETF. Action completed.
- Coordinate and edit a RIPE statistics wish-list paper. Action aborted.
As there has been no input on this there was nothing to edit.
- Produce a document on methods for incremental updates of the RIPE database. Action completed.
3. RIPE NCC
3.1 Introduction by Rob Blokzijl.
3.2 Daniel Karrenberg presented the current status of the RIPE NCC. The NCC personnel were introduced: Anne Lord Network Administrator Starts May 1st Marten Terpstra Network Engineer Starts May 1st Daniel Karrenberg NCC Manager Started April
The NCC have rented 2 rooms from NIKHEF. The office is room number H125 and there is a conference room available (H133). The NCC also have use of the computing facilities at NIKHEF and other support facilities. The computing equipment at the NCC comprises 5 SUN ELC's (2 servers, 3 workstations) together with 3rd party peripherals, which were delivered on 25th March. The public server was operational on 25th March. The NCC computer systems are now fully operational. A separate RIPE NCC Ethernet is connected to the NIKHEF IP router. The public server is reachable via XXX from PSDN and IXI.
NCC activities during the first month of operation included: a revision of RIPE DB software, a pilot interaction information system giving general information and feedback; document store access (Gopher); document store full text search (WAIS); database search. A delegated registry was started. Cooperation with EBONE was initiated. A first draft of a RIPE NCC consistent house style for all documents was established. And lastly, (but not least) the first month activities included preparation for this meeting. The first month credit is due to CWI, NIKHEF - Marten Terpstra and RARE.
3.3 Goals before the next meeting The goals of the NCC follow the "NCC Activity Plan" detailed in ripe-w07. The priorities are not fixed as such and the NCC need guidance from RIPE - specifically the working groups, individuals and the chair. The goals of the NCC are ambitious and in some areas the guidance that the working groups provides is critical.
The RIPE Network Management Database will be updated with regard to the following: new objects; improved update tools; new release of software; list of secondary site; operational distribution service; consistency checking activity; RIPE version of "whois"; exchange with DDN NIC and MERIT.
With regard to network maps,it is necessary to establish a format, possibly hierarchical, for the maps together with a set of tools to be made available for this purpose. It is also necessary to establish a network of maintainers.This was noted as an ambitious task in which guidance from the MAP working group is critical.
For Network Statistics, an SNMP querying tool is being developed following the OPSTAT draft. There is also a need for connectivity monitoring tools which should be installed at strategic sites with the results widely distributed.
There is a need for co-ordination between the NOCs with regard to the statistics gathered. Guidance from the Statistics working group is essential.
The following activities are planned with regard to DNS: Regular hostcount; development of consistency check tool; notification procedures for consistency problems required; Information packagewith regard to leaf domains and high level domains. Secondary service for top level domains and to investigate the need for a Root server.
The need for guidance from the working group is stressed.
A list of NOC`s is planned. This raises questions regarding the maintenance of the list and its distribution.
More generally, Improvements in the support for the RIPE meetings is planned together with the support structure for the Working Groups. Currently work is underway to improve the RIPE document store.
A quarterly reporting period is proposed. The first quarterly report is due in June followed by a September report.
An NCC house style will be drafted for the June report and this will be improved upon for the September report.
There are a number of activities which are NOT in the RIPE NCC activity plan. Specifically, the question of referral. Questions regarding internet services would be referred to suitable providers. End User questions will be forwarded to the appropriate NIC/NOC. It was noted by Tim Dixon that RARE could provide a list of Internet service providers.
A short discussion followed on whether the RIPE NCC should provide a referral service. Concern was expressed that this would quickly grow out of hand. Daniel Karrenberg commented that for the present the RIPE NCC would attempt to resolve user queries, but that he would monitor the situation. It was noted that the information from the queries should be made publically available to ease the work of future referral services.
In summary, a good start has been made. The goals of the NCC are ambitious and require guidance from the working groups. This is critical. It was stressed that the NCC was NOT the panacea to fix everything immediately and magically. However the NCC does plan to operate very much with an "open door" policy, contactable at the following e-mail address: ncc _at_ ripe _dot_ net - will be received by all the NCC staff karrenberg _at_ ripe _dot_ net - will be received by the NCC Manager only. tel: +31 20 592 5065
Marten Terpstra presented the technical installation at the RIPE. This comprised details of how to reach the services provided by the NCC, including the services menu; Whois; File transfer; Gopher and WAIS. Some of the pilot user interfaces were also presented with ann explanation on how to move around the menus and utilise the services.
4. RIPE 4.1 RIPE position paper (introduction)
Rob Blokzijl presented a paper on the relationship between RIPE, RIPE NCC, RARE and RARE RTC. RIPE consists of all RARE members plus others. RIPE has thus a broader coverage than RARE. It is RIPE members that are now implementing the EBONE and will cooperate with the RIPE NCC. In the paper the precise role of the planned OU is not yet finalised. The practical implementation of the OU is therefore unclear. Two options present themselves: if the OU becomes an organization that buys needed services then there will be some similarity to the relationship that exists between the EBONE and the RIPE NCC. Alternatively if the OU is to implement the services internally then the OU will become a service provider - one amonst many.
A position document was drafted during the meeting. An additional topic for consideration is what is implied by "Europeenne" in the RIPE name.
4.2 RIPE vice-chair (introduction)
As Daniel Karrenberg is now employed as the RIPE NCC manager he has resigned from the position of vice-chairman of RIPE. Already it seems that there will be a considerable amount of work to do given that two organisations have approached RIPE regarding support for planned conferences. Specifically IETF would like to hold their first European IETF meeting and Interop will be held in Europe in 1993. For this reason it may be useful to have two vice-chairmen. Rob Blokzijl proposed Antonio Blasco Bonito and Milan Sterba for these positions.
5. Global connectivity
5.1 ROAD (Elise Gerich) Elise Gerich presented the work on Routing and Addressing done by the IAB chartered ROAD group.
ROAD addresses the following problems: IP address space will run out; class B will run out in the next 2-3 years; the size of the routing tables and management issues - address assignment and routing management. The ROAD group are acting as consultants with the Internet as the customer. The output of the ROAD group is intened for the IAB and IETF, the IEPG and CCIRN, the FNC/FEPG, the ANRG and network operators, vendors and developers of software
The future of the Internet is bright. There are numerous activities within the IETF addressing the problem of addressing. CIDR (Classless Interdomain Routing) plans the hierarchical assignment of remaining IP addresses. Addresses are assigned along routing topology considerations. Interdomain routing protocols will agrrgate network numbers along topological assignments. Support is needed for the CIDR plan in the form of standardised technology, operational requirements and an Internet infrastructure. Briefly the CIDR address and routing plan
(i) Sites will get contiguous class C network addresses with the maximal number of initial bits being the same, thus enabling a single prefix for routing (network address w/mask),
(ii) sites adjacent to each other in the routing hierarchy will be assigned out of the same power of two block enabling multiple sites routed with a single routing table entry. Class B exhaustion looks imminent in the next 2-3 years.
The IETF is considering a number of encapsulation issues and issues in migration to CLNP. Possible solutions to IP address exhaustion include using bigger addresses in packet header via CLNP and to make current IP address non global. There is a fundamental problem in the IP address space limits. The themes of the CIDR plan are to develop: better addresses; better use of addresses and to improve address administration.
For those interested in the discussions on the routing and addressing issues a mailing list has been installed named: big-internet _at_ munnari.oz _dot_ au and requests to join the list should be sent to: big-internet-request _at_ munnari.oz _dot_ au.
On another topic, Elise mentioned that according to Olivier Martin the statistics presented earlier on the NSFnet - CERN T1 connection and between Merit and CERN were incorrect. An investigation at Merit revealed that during a reconfiguration an error occurred which has subsequently been rectified.
5.2 PACCOM - Pacific Area Networking (Torben Nielsen)
Torben Nielsen gave a short presentation on Pacific Area Networking, bringing the participants up to date on the latest developments in this area, which was both instructive and interesting.
5.3 Central and Eastern Europe (Milan Sterba)
The RIPE report on ECE countries connectivity has been revisited and new information has been added to the report (see ripe-w09.v4.ps in ~ftp/ripe/docs/ps directory on ftp.ripe.net). The report will be beautified and widely distributed at JNC'92 and at INTEROP spring 92.l
The reports from the Netschool held in Trieste suggested that it would not be useful to plan tutorial work for the forthcoming RIPE Technical days. This is the domain of the Netschool. The RIPE meeting technical days will therefore concentrate on issues related to advanced technical networking.
5.4 NSFnet/SPRINT: ICM project (Steve Goldstein, Bob Collet)
Bob Collet presented a paper on the latest technical developments on the ICM project.
6. Global addressing (Bernhard Stockman)
Bernhard Stockman presented the issues addressed by the ROAD group, indicating that it is necessary to think about IP addressing on a global scale. To this effect a working group has been established. Point 10.7 in these minutes is a report on the discussions of the working group.
7. EBONE 92
Bernhard Stockman gave a presentation of the EBONE initiative and its current status.
The EBONE initiative was started in September 1991 with a proposal for the EBONE-92. A group of interested organizations was formed shortly after that known as the European Consortium of Contributing Organizations (ECCO). The ECCO dedicated the a EBONE Management Committee and an EBONE Action Team to fulfill the implementation of the EBONE. Implementation commenced during the first part of 1992. The topology, the configurations and the operations have since then been specified. The Royal Institute of Technology in Stockholm was accepted as the EBONE Network Operation Center. The EBONE will integrate with other European backbones as the IXI and the EMPP network initiatives and is regarded as complementary to these services. Expansion to EXE countries is foreseen. The Austrian ACOnet has expressed interest in providing an EBS where ECE countries can connect to the EBONE. Increase of bandwidth in the EBONE kernel and between EBONE and the US is needed.
Recently RENATER has joined the EBONE effort. The REANTER contribution will be the EBS installation in Montpellier, part of the 256 Kbps line Montpellier - CERN and the 256 half-circuit Montpellier - London.
The Stockholm - London 256 Kbps line will be delivered May 6. The Amsterdam - CERN upgrade to 512 Kbps is being done now. Madrid - Amsterdam 64 Kbps line will be installed within two weeks. Pisa - Montpellier 64 Kbps line will be upgraded to 128 Kbps.
The EBONE to IXI gateway has been defined by Dutch PTT and NIKHEF and will be implemented as soon as the CEC gives its formal approva to the Dutch PTT.
A routing plan has been completed which gives the detailed explanation of the routing configurations as well as examples on how to configure regional connections to the EBONE and an explanation of the routing protocols to be used within EBONE.
A management and operations plan has been completed which outlines the different levels of responsibilities and the interactions between the EBONE core and the participating networks. The EAT will have the configuration and change control. The EBONE Operations Team (EOT), consisting of one person from each EBS installation and the EBONE NOC, will be responsible for implementing the configurations and changes.
The major EBONE documents are available via anonymous ftp from archive.ripe.net and nic.nordu.net in directory ebone.
8. RARE support for the RIPE NCC (Paul Van Binst)
Paul Van Binst, RARE treasurer, presented the RARE support of the RIPE NCC. Rare acts as a clearing house for the NCC budget of 240 kECU which will cover manpower, equipment and rooms. Funding for the first year of operation of the NCC is provided by the national members of RARE, EARN (who contributed 85k ECU) and Israel.
The OU will play an important role in the future. The budget will be 10-15 MECU per year. The resources will mostly be used to cover line costs for European networks. The next Operational Unit shareholder meeting will be June 5. It is planned to have an interim OU installed three months later.
9.1 RIPE position paper (discussion and decision)
Rob Blokzijl drafted a first version of the position paper, which was distributed at the meeting. The principles contained with the draft were discussed further in a WG session and subsequently restated and agreed in a plenary session with the proposal by Steve Goldstein that the relationship of RIPE to the global Internet should be covered in the document. This was accepted.
A first complete paper is planned to be ready for the September RARE CoA meeting). In addition, the position document will be distributed to all relevant mailing lists.
9.2 RIPE vice-chair (decision)
No other proposal for RIPE vice chairs had been submitted. The meeting unanimously accepted Antonio Blasco Bonito and Milan Sterba as RIPE vice chairs. Before a final decision is possible Blasco and Milan will have to check with their respective home organizations for permission to take on the work.
10. Reports from the working group sessions
10.1 Relationship between Academic & Research networks and Commercial networks - [Glenn Kowack]
Group still working on basic definitions - particularly the definition of what is "commercial traffic" and the catalogue of issues. During the working group discussions, there was a digression into the issue of settlement, which the group noted that they did not want to be dominated by: the focus of the group is on quality discourse. It was noted that there were sensitive issues under discussion. The diaglogue of the group would be continued over the e-mail. The nature of the discussion will be small groups focussing on large issues at upcoming international meetings. Attempts will be made to broaden participation in the discussions. Rob Blokzjil commented that the charter of the group is cataloguing the issue rather than to provide solutions.
10.2 Network Information Services [Nandor Horvath]
10.2.1 Report from the IETF USWG - Joyce Reynolds
During the working group session, Joyce gave a short presentation about the current state of activities in the US in the field of user information services, as follows.
The actual projects of the User Services Area are handled by the creation of Working Groups of which there are currently 8. Two of these groups are newly formed: Internet School Networking and the Internet Anonymous FTP Archives.
IETF User Services Area Working Groups
Directory Information Services Infrastructure (DISI). DISI is chartered to facilitate deployment of X.500 Directory Services on the Internet, by producing "Administrator's Guides". Chair: Chris Welder disi _at_ merit _dot_ edu
Internet Anonymous FTP Archives Working Group (IAFA) is chartered to define a set of recommended standard procedures for the access and administraiton of anonymous ftp archive sites on the Internet. Chair: Peter Deutsch. Alan Emtage iafa _at_ cc.mcgill _dot_ ca
Internet School Networking (ISN). ISN is chartered to facilitate the connection of the United Statess K-12 (Kindergarten-12th Grade) schools, public and private, to promote school networking in general. Chair: John Clement, Art St. George & Connie Stout isn-wg _at_ bitnic.educom _dot_ edu
Network Information Services Infrastructure (NISI). 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. Chair: April Marine, Pat Smith nisi _at_ merit _dot_ edu
NOC-Tool Catalogue Revisions Working Group (NOCTOOL2). "Son of NOCTools" is updating and revising their catalog to assist network mangers in the selection and acquisition of diagnostic and analytic tools for TCP/IP Internets. Chair: Robert Enger & Darren Kinley noctools _at_ merit _dot_ edu
User Documentation (USERDOC2). UserDoc2 is preparing a revised bibliography of on-line and hard copy documents/reference manuals/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 themeselves. Chair: Ellen Hoffman, Lenore Jackson userdoc _at_ merit _dot_ edu
User Glossary (USERGLOSS). User-Gloss is chartered to crate an Internet glossary of networking terms and acronyms for the Internet Community. Chair: Gary Malkin & Tracy LaQuey Parker usergloss _at_ xylogics _dot_ com.
User Services (USWG). 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. Chair: Joyce K Reynolds us-wg _at_ nnsc.nsf _dot_ net
In the discussions that followed within the working group, it was agreed that a unified set of tools were needed to access the service. There are existing tools which need evaluation in order to select an appropriate tools to consistently use. A well known entry point is needed, but it is accepted that we should not put the existing information there, but rather, the information stored, is maintained by those responsible for it and keep only the pointers to such information. Specifically two questions arose: (i) what sort of information is needed? (ii) how to structure the information stored The content and structure of the information to be stored within NIDUS needs to be given careful consideration. A small working group of 4 people will consider this. Geza Turchanyi and Milan Sterba volunteered to be part of the group.
Subsequently in the working group presentation, Joyce introduced a new information service which gives help with getting rfc's. To reach this service send an e-mail message to: rfc-info _at_ ISI _dot_ EDU Subject: getting rfcs help: ways_to_get_rfcs - rfc-editor _at_ isi _dot_ edu is the address to send RFC submissions (postel&jkrey) - IANA _at_ isi _dot_ edu is the address to obtain number assignments (ports, protocols, mibs, PPP, etc) postel&jkrey) - to check whether you are on an rfc mailing list use the following address: rfc-request _at_ nic.ddn _dot_ mil Please use this address to mail all your comments on the service.
10.2.2 Progress of the RIPE activity
The working group adopted the acronym NIDUS for Network Information Discovery for User Sup- port. The e-mail address: nidus-wg _at_ ripe _dot_ net nidus-wg-request _at_ ripe _dot_ net
10.2.3 Acceptable Use statement repository
Information should be sent to the RIPE NCC.
10.3 DNS issues [Francis Dupont]
10.3.1 RIPE DNS Hostcount Marten Terpstra reported on the current DNS hostcount. The hostcount as of April 27th 1992 is 170,050 which represents an increase of 2,111 from March's hostcount. For those interested in the details of the DNS hostcount, a copy of the report is held on the RIPE document store in both ps and ascii format (ripe-i02)
Also in the document store is a bar chart showing the history of the DNS hostcount.
10.3.2 Second Root Name Server in Europe The group discussed whether it is necessary, where to put it, what is the problem priority and what will be the resources usage. Information is needed about the root name server at Stockholm. The NCC will investigate this and the relevant IETF working group will be contacted regarding global requirements.
10.3.3 DNS Debugging Tools Jose Martins reported on the Domain Debugging Tools in Portugal. The NCC stated that it would evaluate the tools and report back.
10.3.4 Reverse Mapping The working group discussions concluded with two recommendations. Firstly every network with RIPE (or more) conectivitiy should have a reverse mapping service and secondly that rejecting connections because reverse mapping service is not available or unworkable, is unfair and sometimes illegal (as in the case of SMTP cf PP software for example)..
10.4 Routing issues [Jean-Michel Jouanigot]
10.4.1 Policy Based Routing Within RIPE document The new version of the document was presented. It was agreed that this version was the final one, and that it should be presented to RIPE for agreement.
Most of the networks present agreed to make use of the new functionality. CERN volunteered to contact its neighbors and implement the new schemes. EBONE plans to make use of the database for routing (see point 10.4.3).
A Network List Compiler (nlc) was presented. nlc generates directly the
routers access-lists from the RIPE database, any kind of input file and a RULE file. A quick overview of the RULE file was given. This software, which is currently used at CERN as a BETA test, will be made available soon.
10.4.2 Routing Policy Description Form
The IETF BGP Working Group "Routing Policy Description Form" was discussed. This form allows networks to publish their routing policy in terms of Autonomous Systems. Very few of the people present agreed to fill this form, very often because it was almost impossible for them to write down their routing policies on Autonomous System basis. The approach suggested by the working group solves the problem on network basis instead of Autonomous systems. It was felt that the two proposals do not have the same "granularity" and that the RIPE database solution allows a better control on the information. However, the working group advises to use BGP wherever possible in order offer two solutions for Policy based routing.
10.4.3 EBONE Routing Plan Principles on the EBONE routing were presented. EBONE will use BGP and the RIPE database. The working group regrets that the EBONE Routing Group Chairman was unable to attend the Working group. Nevertheless, the new "Ebone objects in RIPE database, proposal, P. Lothberg 1992 03 15" was discussed. Most of the fields requested by the proposal are already in the database (net, con, as). It was felt that EBONE dedicated information in the RIPE database was not desirable, and that the working group (net, con, as). It was felt that EBONE dedicated infor-
mation in the RIPE database was not desirable, and that the working group proposal solves the EBONE routing issues. The Working group urge the EBONE Routing Working group to define precisely their requirements if the working group proposal does not fulfill their needs.
10.4.4 Introduction of BGP A few tests have taken place in different sites in Europe, without global coordination. In almost all these tests, BGP is use as a replacement for EGP (BGP II). BGP is now use intensively inside NORDUnet, and EBONE plans to use BGP inside the backbone. Unfortunately, the participants resources do not allow an European Wide Pilot (lack of routers and lines, sometime manpower). NORDunet is asked to publish their results and experiences with BGP. If a site happens to have spare equipments and manpower, and want to perform some tests, it is asked to contact Jean-Michel Jouanigot who will investigate the possibilities. The bgp-pilot list is still operational as bgp-pilot _at_ dxmint.cern _dot_ ch.
10.4.5 AOB The creation of a mailing list was decided: routing-wg _at_ ripe _dot_ net. To subscribe, send your request to routing-wg-request _at_ ripe _dot_ net.
10.5 Network monitoring and statistics gathering [Bernhard Stockman] The session mainly treated the IETF OPSTAT Internet Draft that recently has been submitted and thus is available with anonymous ftp from the the IETF Internet-Drafts archives. A presentation was made of the content of the draft. The mailing list for the OPSTAT work is: oswg-l _at_ wugate.wustl _dot_ edu and request to be added should be sent to: oswg-l-request _at_ wugate.wustl _dot_ edu.
As an example of link utilisation statistics, Joseph Michl presented an histogram of peak-distributions for a BelWue to DFN/WIN link. It was concluded that there is a need for more work in this area to define characterisation of link utilisations.
Peter Houlder from UKnet presented his experience with the Monster NMS-tool. UKnet are using 10 minutes polling periods. The autoconfiguration part of Monster was appreciated. By use of the polling logs, events and plots are extracted into traffic diagrams.
10.6 Global Addressing [Bernhard Stockman] The working group session was discussing the content of a proposal for a global addressing scheme. The need for such a scheme is based on the current Internet address space exhaustion and the growth of the routing tables which creates management problems as well as hard and software problems in Internet routers.
The proposal shall contain the basic ideas on a partitioning of the class C address space and the criterias used. There is a need for a specification of how the distributed management of a partitioned address space shall be implemented. Problems like holes in the address space, multi-
homed networks and networks that move between service providers shall be covered. The timeframe within the proposed solution is valid should be estimated. As there will be allocation to the current assignment method there is a need to set aside some part of the address space for this and to define the amount of space available for partitioned assignment.
It was believed that initially only a smaller fraction of the class C space should be allocated for partitioned assignment until more experience has been achieved. Once such experience is there bigger chunks could be allocated.
10.7 Network Maps [Hank Nussbacher] Hank Nussbacher could not attend this meeting. No Network Maps session was arranged. Don Stikvoort volunteered to update the RIPE maps. Updates should be sent to stikvoort _at_ surfnet _dot_ nl. It was also noted that the RIPE NCC is chartered to provide network maps. There are two target groups, administrative people having need for overviews in their planning and technical people having need for more detailed maps. Daniel Karrenberg expressed the need of having general mapping formats. One issue for the mapping WG could thus be to define how maps should look like and which tools should be used to produce the maps.
10.8 European connectivity [Milan Sterba] The current situation of ACONET, its connectivity to Ebone and its support to ECE countries connectivity have been shortly presented.The project of hungarian academic backbone has also been presented.
Rob Blokzijl has presented the current situation in the xSU. There is an urgent need for IP connectivity at least to Europe. A creation of a single higher speed connection to some place in Western Europe, shared by all those involved in networking in xSU. All parties involved in networking withinxSU and with xSU are invited to cooperate closely to achieve a better connectivity and a better level of network services.
The results of the PC router developpments of the COPERNICUS project lead jointly between FESnet Czechoslovakia and INRIA France have beenpresented by Jiri Kaspar and Ivo Cernohlavek. A cheap PC router solution (max 1KECU + PC) allowing TCP/IP routing over Ethernet and X25 links(up to 64 kbit/s) based on OST, RCE and Synaptel X25 boards has been presented by Jiri Kaspar (kaspar _at_ cs.felk.cvut _dot_ cs). A similar workoriented on synchronous PPP on a own-designed board is being finished by Ivo Cernohlavek (firstname.lastname@example.org). These results of COPERNICUS will be used in the implementation of the Czechoslovak academic backbone.
Coordination among the academic networking initiatives of ECE countries progresses. A new coordination meeting (the 1st has been held in Prague in February 1992) will be held in Innsbruck on the occasion of JNC'92. progresses. A new coordination meeting (the 1st has been held in Prague in February 1992) will be held in Innsbruck on the occasion of JNC'92.
10.9 RIPE Database [Wilfried Woeber] The RIPE database status report was presented by Daniel Karrenberg. The database of RIPE network management information continues to grow. As of April 1992 the total number of objects was 6967. In June 1991 this figure was 3,168. Those interested in the details of the report should consult the RIPE document store where the report is held in both ps and ascii formats (ripe-i01).
The RIPE-DB Working Group (23 Attendees) dealt with the following issues:
a) The more traditional Database-Objects The group reviewed the current status of the Link, MHS, Ebone/Router, AS objects. Action items were defined for Ruediger Volk and Daniel Karrenberg to update and recirculate current proposals and recommendations to get the group moving again.
A. Blasco Bonito initiated a discussion whether it is feasable and useful to add more items to existing objects. He was invited to circulate a proposal and some examples of the extensions used in Italy on the RIPE-List.
A general discussion evolved around the issue, whether it is preferable to add more information to the RIPE-DB or to employ some new Record-Types, defined for the DNS recently. Main issues in the discussion being whether this is to be considered for a long-term development or a short-term solution of urgent needs. One of the aspects was that the RIPE-DB is available and easily expandable, while the use of the DNS would require to employ updated software releases in all the systems involved. On the other hand the DNS software is generally available at almost all environments connected to the internet. The group did not develop a common point of view or rec- ommendation. Open for further discussion and input.
Daniel Karrenberg expanded on his report on the current state of the RIPE-DB Software. Major facts are that the software is freely available on request. However, the RIPE-NCC is tasked to keep track of "secondaries". Everyone running a "public" whois server is strongly urged to inform Daniel asap.
Updates for the RIPE-DB are to be sent to ripe-dbm _at_ ripe _dot_ net The new release of the software will include inproved consistency checking and mechanisms for regularly distributing updates to secondaries. Availability of updates can either be announced through email or sent automatically by FTP.
b) Recent requests and current situation with regard to X.400,X.500,etc. With regard to the COSINE, IXI, CONCISE, PARADISE, et.al. issues the lack of representatives was disappointing. Thus, although future collaboration was felt to be desirable, the group could not technically discuss the open issues. The attendees agreed that a "formal invitation" should be forwarded to the projects involved to be represented at the next RIPE-Meeting in Paris.
In the meantime everybody was invited to follow up their personal contacts to get the discussion going.
The RIPE-DB-WG decided, that the information exchange and discussions in relation to the group's work shall use the RIPE-List.
11. Date, place and time of next meeting
Next meeting will be September 30 - October 2. France Telecom offered to host the meeting in Paris. The meeting accepted the offer.
At the Paris meeting there will be technical sessions. One suggestion for this program is to have a demo of the INRIA developed cheap PC-board based routers. Ideas on the technical program should be sent to the RIPE NCC which will coordinate this.
The first 1993 meeting will be January 25 - 27. The venue, Prague was proposed and has since been confimed.
12.1 GARR Network Information Service. Antonio Blasco Bonito presented the new GARR information service. The need for such a service has been recognized since GARR was first implemented three years ago.
The information services include: registry of organizations; people; IP networks and e-mail domains; top level DNS server; central GARR working group's archive and archives of information documents.
The manpower to provide the service currently accounts around 2-3 man/years.
The hardware consists of a Sun server and four 486/unix workstations.
The GARR Network Information Service can be reached via email on the address: nis _at_ nis.garr _dot_ it.
The ftp archive is accessible on: ftp.nis.garr.it.
Information on the service is can be requested via email on address: info _at_ nis.garr _dot_ it.
12.2 On the DIGI work Dave Morton, ECRC, presented current work within Deutsche Interessen Gemeinschaft Internet (DIGI). DIGI has a charter similar to ISOC. At a meeting in December 1991 it was a 50/50 percent mixed participation between commercial and academic organizations. DIGI is currently negotiating with its solicitors on making DIGI a legal entity.
A meeting is planned in November 9-12th 1992 in Munich where colleagues from various bodies in Europe and US will be invited. Tutorials are planned for the first two days of the meeting (9-10th November).
Rob Blokzijl thanked the participants and declared the 12:th RIPE meeting closed.
APPENDIX - ACTION LIST
Ruediger Volk and Francis Dupont
- Collect DNS tools and make available to public via NCC Archives
Tony Bates and Marten Terpstra:
- Produce a document on experiences of running IP/IXI.
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. Input on the contents of the document is requested.
- Make a new distribution of the database tools available. (Note - the work has been started.) - Coordination of vendor presentations at the next RIPE technical session. - Send the address of the ISO sub-committee that handles ISO-3166 work. Action done.
- To store and make available to public DNS tools. - To install mailing list on RIPE machines for each working group. - To archive routing recommendations in NCC document store. - Collect information regarding the Second Root Name Server. The relevant IETF working group will be contacted regarding global requirements. - To evaluate Domain Debugging Tools.
Hank Nussbacher and Don Stikvoort:
- Write a charter for the RIPE mapping working group.
- Distribute the draft RIPE statistics document to the ripe-org mailing list for further comments.
- Produce a document on methods for incremental updates of the RIPE database.
- Send routing recommendations to RIPE for approval.
- Make available NLC software via anonymous FTP (via NCC).
- Discussions regarding the Routing Policy Description Form. Discuss this with Jessica Yu.
Milan Sterba and Geza Turchanyi:
- Evaluate and consider nature of information for NIDUS.