.po 1.4i .sz 11 .nf \fBNORDUNET CLNS PROJECT\fR 6.10.1990 \fBGeneral Overview\fR .fi CLNS will soon become an important service within the RARE community. The reasons for this include introduction of CLNS based DECnet Phase V and increasing availability of both public domain and commercial CLNS capable OSI applications. Also in LAN environments (to which most computers are attached), CLNS seems to be the manufacture network service of choise. The goal of the CLNS Project is to gain experience of CLNS networking using real NSAP addressing. As the result, the the NORDUnet backbone and national backbones should become common carriers for both DECnet Phase V and other CLNS applications. The project consists of (1) producing recommendations for the Nordunet community, (2) a CLNS pilot service, and (3) a DECnet Phase V migration pilot. It will also follow the progress in the area of transport service bridging, eg. by monitoring the work done by RARE. \fBDeliverables and Time Plan\fR The CLNS Project will last for one year starting at the beginning of October 1990. The different parts of the project will be covered in more detail below. 1. \fBRecommendations\fR The recommendations will suggest: (a) An NSAP format suitable for CLNS taking into account the requirements of IS-IS and IDRP routing as well as interworking with CONS. The NSAP recommendation will, of course, affect only the local part of the DSP left unspecified by the national ISO body. (b) A mapping of the NSAP format (administrative authorities, routing domains, and areas) into the Nordunet network architecture. The mapping should take into account the possible requirements of point (3) below. (c) A DECnet Phase V transition plan. These recommendations can be delivered by the end of April 1991. 2. \fBCLNS Pilot\fR The CLNS pilot will consist: (a) A ISO IP IS-IS routing pilot which itself has two consequtive stages: - ISO IGRP stage (from October 1990 until April 1991), and - ISO IS-IS stage (from April 1991 until the end of the project). The deliverable of this part includes an evaluation on the scalability of IS-IS routing protocol and a recommendation on how interdomain routing should be applied. Note, that at the projected beginning of the second stage, the number of available ISO IS-IS implementations may be limited or even zero. The stage cannot start in larger scale until at least the Cisco ISO IS-IS implementation is available. (b) Interoperability tests between Digital and Cisco ISO IS-IS implementations over different media. This can start earliest in April 1991 (when both Cisco and Digital ISO IS-IS implementations are available) and will last until end of the Project. The deliverable is a report describing the possible incompatibilities and guidelines for configuring the routers for interoperability. (c) ES-IS interoperability tests between Cisco (and possibly also Digital) ES-IS implementation from the IS side and various available ISO IP products (Digital Phase V and VOTS 2.0, SunLink OSI, Retix 400, and others) from the ES side. The phase can start in October 1990 and will last until the end of the project. The deliverables include possible problem reports as well as a guide for configuring various ISO IP end systems for ES-IS routing. (d) Participation in RARE and the US Internet CLNS pilots. 3. \fBDECnet Phase V Transition Pilot\fR This project will only be concerned with the Network layer (and below) of DECnet/OSI (DECnet Phase V). Other topics which are important for DECnet/OSI, e.g. the name-server/space, are not considered important in THIS project and neither are the details of the transition phase from DECnet Phase IV to DECnet/OSI. The requirements put on NSAP structure, topology etc. by DECnet/OSI will, however, be considered. Most of the stages, aims and results described in the General Overview and under 1 and 2 above apply also to DECnet/OSI, or explicitly involve DECnet/OSI systems. DECnet/OSI is to be seen, in so far as possible, as a part of a general CLNS network. Some unique issues are: (a) Summarize implications on CLNS topology and NSAP structure from DECnet/OSI, both during the transition phase and in the long term. (b) Investigate possible restrictions caused by the DECnet/OSI Nameservice. The Nameservice as such is NOT a part of this project. (c) Consider the alternatives available for connection to HEPnet and SPAN both during DECnet Phase IV to DECnet/OSI transition and in the long term. Current systems can not be used and the availablity and suitability of possible equipment must be determined. (d) Evaluate impact on NORDUnet CLNS from current DECnet/OSI migration plans within HEPnet and SPAN Given some recent information, parts of the DECnet/OSI software and equipment may be delayed until the middle of 1991. Initially, it may only be possible to use dedicated DECnet/OSI routers (IS-IS) and Ultrix ES. The interoperability test between DECnet/OSI and other CLNS equipment described in section 2 above may be delayed due to this. \fBInitial Pilot Participants\fR Finland - FUNET, Juha Heinanen (SunLink OSI, Retix 400, Cisco), Jyrki Soini (DECnet Phase V and VOTS 2.0 end node) - Telecom Finland, Juha Heinanen, DataNet backbone - Digital Finland participation expected via DataNet Norway - UNINETT, Havaard Eidnes (DECnet Phase V end node, SunLink OSI) Sweden - SUNET, Goeran Bengtson (DECnet Phase V end node and possibly also a router, SunLink OSI) - Digital Sweden, Krister Larsson (DECnet Phase V end node and router) Other Nordunet sites are welcome to join at a later stage. \fBBudget\fR The cost of the project is 160KSEK. This sum is allocated for project management, travel, and producing the documents of the above mentioned deliverables as follows: (a) project management 30.000 SEK (85 hours at 350 SEK) (b) travel 60.000 SEK (3 project meetings, 4 persons each + 1 low cost trip to CERN by 1 person) (c) document production 70.000 SEK (two reports, 85 and 115 hours at 350 SEK) National networking organizations are expected to contribute the work and equipment needed by the pilots.