- ripe-list - mailman.ripe.net

MICE - not quite there yet...
by Jon Crowcroft 29 Sep '92

29 Sep '92

-------- the message i sent about multimedia for europe was a bit premature - sorry, enthusiasm got the better of me. the project is not yet funded... you will certainly know as soon as we find out! note, also, that the protocol stack is subject to much study (as one would expect for continous media, there are no firm standaards yet, so some investigation is needed before committing to any particular technmology...an isochronous service is offered byu some x.25 switch manufactuerers, for instance (e.g. netcomm)). sorry to over-raise hopes... j. --------------- To: matsb(a)sics.se Cc: Piet.Beertema(a)mcsun.eu.net, wallace(a)rare.nl (Bruce Wallace), cpmu(a)rare.nl, davies(a)rare.nl, kalin(a)rare.nl, roll(a)sbd.stupi.se, ripe(a)ripe.net Subject: Re: Mis-information In-Reply-To: Your message of "Tue, 29 Sep 92 12:48:24 BST." <9209291148.AA03177(a)asia.sics.se> Date: Tue, 29 Sep 92 13:37:43 +0100 From: Jon Crowcroft <J.Crowcroft(a)cs.ucl.ac.uk> >Waht we do need is risk capital from tax payers to show new >applications so taht the service providers will invest in higher >bandwidth network solutions for mulitmedia mail and video >conferencing. >Step 1 34 Mbps addtional bandwidth support to Ebone, maybe PTT telcomm >is interested to prowide this for the same amount of money? >Step 2 Separate testbed to try new concepts and protocol structures yes - we will need just these two steps to finalise work on the MICE project in late 1993. However, to start this project (from the day after tomorrow:-), we need: For those of you not aware, the EC has funded MICE (Multimedia International Conferencing for the Ec) to coordinate pilot services for audio, video, audio and shared whiteboard/editors between any interested EC and EFTA countries....at least France, Germany, Netherlands and Sweden + Norway are involved already. We will be using existing tools (e.g. vat, dvc, inria, videoconf,) for workstation conferencing and H.261 CODECs with packetizers and any other technology to link video conferencing roms facilities at Research and Educational Facilities. (A Packetiser is usually a sparcstation with hsi board). typically, this will entail IP (since much of the video/audio technology comes from the US< just like good networking technology), to carry packet video and audio - we hope to work with RIPE/RARE and with router and switch vendors to influence the production pof resource guarantees in ther network devices. End systems are already fine for audio, and shared applications in the Ebone environment. Video has not been extensively tried. If you are worred about being swamped with traffic, let me assure you that each user nation *must* have full approval end to end... obviously, the multicast facilities of the Mbone will be essential in optimising traffic (will EMPB provide multicast?). jon and just to agree on the sign-off: > Who would want to participate in an IP *pilot*, when > we have a globe-wide, reliable, and well-managed IP > *service* right behind our keyboards??? ------- End of Forwarded Message

1 0

The Complete EBONE Up and Running !!!
by boss＠ebone.net 29 Sep '92

29 Sep '92

To the European IP community, This is to inform you that today, Sunday 26 September 1992, the final link of the initially defined EBONE, the London-Montpellier link, was put in place. The EBONE is now complete as a resilient pan-European IP backbone. September 1, 1991, after an initiative by Kees Neggers, a small group of people gathered for the first time, in an office building outside Amsterdam, to see if it could be possible to build a pan-European IP backbone. Today, 390 days later, the full backbone is up and running. Without all the voluntary efforts put in this would never have been possible. As a result we today see one of the worlds technically most advanced backbones in Europe. I must admit I have never seen a group with so much expertise gathered around the table and with so much enthusiasm in making this knowledge serve the European networking user community. I will by this give my deep gratitude to all those people involved in the fulfillment of the first pan-European IP backbone. To make it short: WE MADE IT!!! ======================== There is some bottles of champaign waiting in Paris :-) See you there all of you, Bernhard Stockman. Chair of the EBONE Action Team.

6 5

Re: Status of connectivity to xSU by US Federal nets
by Arnold Nipper 29 Sep '92

29 Sep '92

>PS: Countries whose nets are announced on the NSFNET Backbone are listed in >/statistics/nets.by.country, available by anonymous ftp from nis.nsf.net. -s /statistics/nsfnet/nets.by.country Arnold Nipper XLINK

1 0

Status of connectivity to xSU by US Federal nets
by Stephen Wolff 29 Sep '92

29 Sep '92

On behalf of the R&E community, NSF sought and in June, 1991 received a letter from the Department of Commerce authorizing IP connectivity to the so-called Eastern bloc. In a second letter from Commerce in January, 1992, the authorization was reaffirmed and clarified; i.e., that the concern was export control, **not** the connectivity itself. Accordingly, Poland, Czechoslovakia, and Hungary, have been now for some time announced on the NSFNET Backbone; somewhat more recently, Estonia is reachable as are (potentially) the other Baltic states. Although connectivity to the xSU is covered by the Commerce letters to NSF, at the request of the Federal Networking Council (FNC) the NSF agreed not to carry such traffic on the NSFNET Backbone. This concession was nugatory until two recent events: a network in the xSU is now being announced by the CIX, so that it can (and does) exchange traffic with non-Federal US nets, and Federal mission agencies are establishing what amount to point-to-point links from specific sites in the xSU to specific US computational resources. At the FNC meeting on Thursday, the 24th September, the NSF therefore asked the FNC to reconsider its position, since a side effect of the growing connectivity (which is either "non-R&E" or is between unilaterally-selected scientists and sites in the US and the xSU) is to disenfranchise a large fraction of the US R&E community whose aspirations for unrestricted scientific and educational collaboration with their peers and colleagues in the xSU NSF fully endorses. -s PS: Countries whose nets are announced on the NSFNET Backbone are listed in /statistics/nets.by.country, available by anonymous ftp from nis.nsf.net. -s

1 0

Re: Mis-information
by bob＠informatics.rutherford.ac.uk 29 Sep '92

29 Sep '92

> >This means that the IP-community of Europe will probably have > >a zero influence on the network, that are supposed to serve our > >community. And that the CPMU totally ignores the skills of the > >Ripe community. > > This will only be true if none of the Ripe member networks > paticipate in the IP pilot. We (ie JANET/JIPS) intend to participate in the EMPB IP pilot. The reason we're doing it is that we believe it's important to have high-quality networks available for our users, and piloting is a necessary part of the quality-assurance procedure. This attitude is no different to our attitude to Ebone '92 -- we've been happy to participate, and be active in the management and operational work to get things sorted out. It's been a rewarding experience for us, our users, and (I'd suggest) other members of the Ebone consortium. I find it hard to see much difference between doing this and doing the same for EMPB -- in both cases we see a reward for work invested. I'd hope other networks connecting to EMPB participate in the pilot as well. Otherwise there won't be much to pilot :-) > >We have to pay twice, as the skills of the coumunity are not > >used, and we have to pay the consultants on our taxes/ charges. > > You might have to pay for the consultants by your taxes if your > organisation is part of COSINE, but NOT in charges for the EMPB. It seems to me that it's a bit dangerous to start suggesting that one IP network is a burden on the taxpayer, whilst another isn't. As Bruce points out, EMPB charges aren't. Even a cursory inspection suggests that substantial bits of the Ebone '92 infrastructure are publically funded, at least originally. Nothing particularly wrong with that -- they've been put to good use -- but it seems fairly silly to start this particular line of discussion at this particular time. Bob Day

1 0

Re: Mis-information
by Peter Lothberg 29 Sep '92

29 Sep '92

> This means that the IP-community of Europe will probably have > a zero influence on the network, that are supposed to serve our > community. And that the CPMU totally ignores the skills of the > Ripe community. > This will only be true if none of the Ripe member networks > paticipate in the IP pilot. > Who would want to participate in an IP *pilot*, when > we have a globe-wide, reliable, and well-managed IP > *service* right behind our keyboards??? It's even worse, the EMPB has to have it's ovn relationship to EVERY ip network in the wolrd, as they do not cooperate with -GIX- type of concortias like the Ebone. -Peter

1 0

Mis-information
by wallace＠rare.nl 29 Sep '92

29 Sep '92

Dear all , A correction to the "rumours" that are propogated below, from the person involved in the matter. _____________________________________________________________ >Date: Sun, 27th Sep 92 21:14:30 MET >From: Peter Lothberg <roll(a)bsd.stupi.se> >To: ripe(a)rip.net >Subject: What are the COSINE/CPMU in the Rare officie doing? >Some days ago we all received a long document that looked like >this; >Note the document has bee recieved September 8, 1992. >______________________________________________________________ > RARE and RIPE / RIPE - NCC > RIPE: > RIPE should stay a technical IP coordinating body within >RARE, but having much more interaction with the RARE management, >technical and executive bodies (CoA, RTC, REC, Raesec) than in >the past > according to it's charter, RIPE should be serving wider >population than RARE membership > one should obsercve gradual merger (or at least very close >cooperation between identical or related Working Groups of RARE >and RIPE, leading to a coordibated management of technical >activities > Ripe should support establishment and utilisation of the >EMPB, particularly in the pilot phase, where its expertise is >crucially needed >I have learned through releiable sources that the EMPB?PTT- >Telecom is planning to subcontract "IP network Consulting" for >EMPB to Softlabs. This is totally incorrect! PTT-Telecom have not contracted anyone for "IP network Consulting". Softlab are currently contracted to the CPMU for the COSINE project P4- "Network Management in the Context of concantenated Networks". They have been asked to provide consultancy in the above topic for the EMPB.(By the CPMU). >Another strange thing seems to be that the receiver(What's >that?)was " Bruce Wallace at PTT Telecom" I thought he was with >the CPMU? You are quite correct about this point I (Bruce Wallace) am a member of the CPMU and not a PTT-Telecom employee. (I suggest you stop using this "reliable source" as he(or she) is plainly not reliable! >This means that the IP-community of Europe will probably have >a zero influence on the network, that are supposed to serve our >community. And that the CPMU totally ignores the skills of the >Ripe community. This will only be true if none of the Ripe member networks paticipate in the IP pilot. >We have to pay twice, as the skills of the coumunity are not >used, and we have to pay the consultants on our taxes/ charges. You might have to pay for the consultants by your taxes if your organisation is part of COSINE, but NOT in charges for the EMPB. >I might be wrong, if someone has better knowledge about this, >please speak up. >Peter. I have spoken up and I hope cleared up the mistakes. Regards Bruce.

4 3

Re: Mis-information
by Peter Lothberg 29 Sep '92

29 Sep '92

>I have spoken up and I hope cleared up the mistakes. >Regards >Bruce. Thanks Bruce, My problem, and I guess that that goes for many in the IP networking world of the planet Earth are not used to the 'big-brother' approach. The quality of the service, seen by the enduser (and the networks are to support the endusers) are the sum of all involved organisations, that his traffic may traverese. A well managed, but SECRET cloud in the middle, that tryes to establish itsef without talking to ALL of the outside world, creates problems, real of imagined. As you seem to be informed, I would like to ask you to submitt to the Ripe malinglist all relevant documents, and the in deepth technical plan for the EMPB IP service, and how that interropetartes with other planned Unisource IP networking activities. -Peter

1 0

RIPE meeting - Agenda (latest version)
by k13＠nikhef.nl 29 Sep '92

29 Sep '92

ripe-a-13 R I P E Agenda of the 13th RIPE meeting =============================== September 30 - October 2, 1992 - - - - - - - - - - - W E D N E S D A Y 14:00 h - - - - - - - - - - - - - 1. Opening (R Blokzijl) 15 min o welcome o apologies o introduction o meeting secretary o dinner for Thursday evening o approval of the agenda o papers tabled o organisation of the meeting 3. Minutes of the last meeting 15 min o approval of the minutes o action list 2. RIPE position paper (R Blokzijl) 15 min o formal approval of the last draft 3. RIPE NCC (D Karrenberg) 60 min o report o the future of the RIPE NCC: - organisational position - budget - financing 4. RIPE and RARE 60 min o report from the RARE CoA, September 24-25. (R Blokzijl) o the RARE technical program (RTC members) An introduction of the RARE technical program, with an emphasis on IP networking. 5. Global Internet traffic measurement (T Nielsen) 30 min o Torben will present a proposal to do a global Internet traffic measurement. The proposal is that (some of) the RIPE networks participate in an effort to measure global IP traffic streams. 6. Global Address Assignement (D Karrenberg) 15 min 7. Delegated Registry inside Europe (D Karrenberg) 15 min o coordination between local NIC/NIS and the RIPE NCC 8. Networking in Russia (M Popov) 15 min o status, including international connectivity o future plans - - - - - - - - - - - - - T H U R S D A Y 09:00 h - - - - - - - - - - - - - - 9. EBONE (P Jones, B Stockman) 60 min o status report o plans for EBONE 93 10. EMPB IP services (PTT Telecom NL) 30 min o EMPB is going to introduce IP on it's network, EMPB will be connected to some of the RIPE coordinated networks. Issues related to interworking between EMPB/IP and the European Internet will be presented and discussed. 11. Audio conferencing (D Karrenberg) 15 min o introduction o proposal for an MBONE * * * * * * * * * * * * * * * * * * * * * * * The following sessions are meetings of - the RIPE working groups - special interest groups They take place with a degree of parallelism. A. Relationship between Academic & Research networks and Commercial networks [Glenn Kowack] B. Network information services [Nandor Horvath] C. DNS issues [Francis Dupont] D. European connectivity [Milan Sterba] E. RIPE Database [Wilfried Woeber] F. IP routing in Europe [Jean-Michel Jouanigot] G. Audio Conferencing [Daniel Karrenberg] H. Global Internet traffic measurement [Torben Nielsen] I. Delegated registry coordination [Daniel Karrenberg] J. EBS managers meeting (closed) [Phil Jones] K. EBONE Germany meeting [Stefan Fassbender] - - - - - - - - - - - - - - F R I D A Y 09:00 h - - - - - - - - - - - - - - 12. Reports from the working groups. 120 min 13. Coordination of the RIPE and RARE technical programs. 30 min o for decission 14. Global Internet traffic measurement (T Nielsen) 30 min o proposal o decission 15. Date, place and time of next meetings 15 min o January 25 - 27, Prague 16. A.O.B. 15 min 17 Closing RIPE working group: European IP connectivity Chairman: Milan Sterba -------------------------------------------- Here is the first draft for our agenda at the next RIPE meeting in Paris. I consider a meeting of this WG during the Paris meeting as usefull and suppose that we will need two 1.5 hour blocks to cover the topics. Thanks for comments and suggestions. * CEEC report revisited (discussion -> new report before end oct.) * Ebone/non-Ebone coutries/networks relationship (clarification) (this related to a more general problem of AUP's in different european networks) * CEEC connectivity scheme optimization (with respect to existing and realistically planned connections, eg. IXI extensions, recent Ebone evolutions etc.) discussion -> recommendation * commercial use of international links in poorely connected countries (current state and possibilities) <-> (Glenn's WG) * non-connected coutries - state of the art and what can we do * overview of actions to support networking in CEEC countries (joint research, network schools, conferences, direct help, etc.) summary -> NIDUS WG * regional mapping process coordination for CEEC (regional mapping authority appointment) * towards a global network architecture in Europe in the environment of multiple IP providers (regional NAP's, routing servers and arbiters) (BOF - is this scheme suitable for Europe ?) Of course there is one major theme in European connectivity, which is the EMPB IP Pilot/Ebone and OU/Ebone relationships. But I suppose this topic will be covered in plenary. RIPE working group: IP routing in Europe Chairman: Jean-Michel Jouanigot ---------------------------------------------------------------------------- 1- Policy Based Routing Within RIPE 1.1- Presentation and status of nlc 1.2- Progress of the Policy based routing Pilot 1.3- Policy based routing tags collection (draft) 2- EBONE routing 2.1- Status of the EBONE routing 2.2- Backup to the US 2.3- Ebone routing and RIPE database (draft) 3- Routing Policy Description Form (Jessica Yu) RIPE working group: whois database Chairman: Wilfried Woeber ---------------------------------------------------------------------------- DB-WG 1. RIPE-DB Operational Issues 1.1 - change of DB format by NIC (nicnets.db) issue raised by Ruediger Volk, Marten Terpstra 1.2 - addition of, or "nonstandard use" of fields issue raised by Jean-Michel Jouanigot, Marten Terpstra DB-WG 2. RIPE-DB Technical Issues 2.1 - definition of and agreeement on Autonomous System Object issue raised by Daniel Karrenberg >>input solicited from Ruediger Volk, Daniel Karrenberg 2.2 - proposal to add field *tech-l/*tl to objects where useful proposal by Andreas Schachtner 2.3 - modification of *in field to take care of supernetting proposal by Andreas Schachtner 2.4 - review of importance of Domain Object, superceded by RFC1183 issue raised by Daniel Karrenberg 2.5 - discussion on feasibility of general purpose extensions to DB >>input (example for GARR usage) solicited from A. Blasco Bonito DB-WG 3. RIPE-DB Organizational and Legal Issues 3.1 - ownership and "acceptable use" of RIPE-DB 3.2 - review/update of ripe-w02 describing the database objects 3.3 - coordination and future liaison with X.400/X.500/MHS activities

1 0

Notes on some TCP/IP WAN activities in xSU
by Oleg Tabarovsky 28 Sep '92

28 Sep '92

Notes on some TCP/IP WAN activities in xSU ------------------------------------------ Activities in the field of TCP/IP networking include SU domain coordination, assignment of IP numbers to ex-Soviet organizations, development and operation of IP-network in SU. SU domain coordination and assignment of IP numbers are carried out by RelTeam Ltd. private company. xSU/RU NIC is hosted by RelTeam Ltd., it assigns IP numbers on behalf of RIPE NCC. RelTeam Ltd. is also the main developer of all networking software in EUnet/RELCOM network, and the coordinator of IP activities in EUnet/RELCOM. Here follow some notes on current state and future of TCP/IP networking in EUnet/RELCOM network operated by Relcom Corp. Currently EUnet/RELCOM network connects approximately 3000 sites all over xSU. EUnet/RELCOM includes approximately 70 regional nodes (backbones) in all former USSR republics and major cities. Currently most links between nodes and sites in RELCOM are UUCP links. International link now is UUCP connection to EUNet via host techno.fuug.fi in Finland. The growth of the network, significant delays because of UUCP links, inability to provide net services going beyond e-mail forced us to begin the turn from UUCP to TCP/IP network. The main problems we are facing are lack of at least medium speed digital channels (64K), of appropriate routers (cisco's are still under COCOM restrictions for us) and of properly trained staff at every EUnet/RELCOM node. Nevertheless, we try to solve all this problems (probably with the exception of the last one :-)). Starting from the end of 1991 the kernel of xSU IP network has begun to grow. Currently pilot IP network connects 11 LANs in Moscow, Moscow region and Barnaul (Altai). The links are mostly voice grade leased lines equipped with V32bis modems and dialup links (with permanent connection) via high quality phone service called ISKRA (former communist's phone network). ISKRA phone network connects almost all major xSU cities and its quality permits reliable V32bis modem connections. Currently Moscow backbone consists of 3 routers at EUnet/RELCOM main hub (KIAE premises), ISKRA POP and Moscow Long Distance Exchange. All other networks which form current IP network kernel are connected to these routers. Plans call for the establishment of international IP link to EUnet cisco in Amsterdam through the router at Moscow Long Distance Echange in just a few weeks. Voice grade leased line which will form the link is going to be equipped with ZyXEL modems at both ends. That will allow to exchange data at speed up to 16800 bps (uncompressed). During the first stage of IP network development in xSU we plan to connect all major EUnet/RELCOM nodes via voice grade leased lines and ISKRA dialup links. In the near future (until the end of 1992) we hope to establish up to 20 IP links via leased lines and several other links via ISKRA, these links will form the global IP network connecting at least 50 LANs. Regional backbones will in their turn form intraregional IP networks in such areas as Siberia, Ural, Altai, St.Petersburg, Ukraine, Far East and others. Now few technical details on currently used software and hardware. Lack of appropriate routers forced us to develop our own solutions. First of all we developed CSLIP software which is capable to use several physical links for one network interface. Such software allows us to achieve acceptable bandwidth (for example ~6KB/sec FTP for ASCII files) in the absence of medium and high speed links, and provide good backup in case if one of the links fails. We plan to develop in the nearest future the same software for PPP protocol. Routing software currently used is "routed" but it has become unacceptable due to significant growth of RIP traffic over the links. All major routers (Moscow backbone routers first) will be upgraded to use "gated" software (the latest "gated-alpha" and gated-3.0 when it will be released). Using gated will allow to use BGP-3 between major border routers, highly reducing routing traffic. Using BGP-3 will also allow simple routing between xSU AS and Amsterdam cisco. Future plans based on utilizing gated software includes switching to OSPF as IGRP. Until that time networks will still speak RIP to backbone routers. Router's built-in software for "firewall-alike" functionality, IP accounting and priority queues is also under development and will be installed soon. Router's hardware currently consists of 386/33 (486/50 planned), 8/16 channel DigiBoards and Ethernet card(s). We also plan to install appropriate 64K interface. Contacts for EUnet/RELCOM & RelTeam Ltd.: Nickolay Saukh - nms(a)ussr.EU.net Valery Bardin - fox(a)ussr.EU.net Oleg Tabarovsky - olg(a)ussr.EU.net

1 0