Draft Report on the
 
                    WORKSHOP ON CO/CL INTERWORKING
 
 
                       Corporation for National 
                          Research Initiatives
 
                           July 24-26, 1990
 
 
                    Reported by Workshop Co-chairs
 
                Phillip Gross (CNRI) and Les Clyne (JNT)
 
 
 
1. Introduction
 
On July 24-26, 1990, an invited panel met at the Corporation for 
National Research Initiatives in Reston Virginia to consider the
issues involved with interworking between protocol stacks based 
on Connection-mode Network Service (CONS, or CO) and 
Connectionless-mode Network Service (CLNS, or CL).  The main example 
of a CO stack is OSI TP0 over X.25.   Examples of CL protocol
stacks include OSI TP4 over CLNP and TCP over IP. 
 
The workshop was convened at the direction of RARE and the U.S.
Federal Networking Council (FNC).  The meeting was organized and
co-chaired by Les Clyne (UK Joint Network Team) and Phillip Gross
(Corporation for National Research Initiatives).  An electronic
mailing list was established for use by both attendees and a wider
audience of experts.
 
This report gives an overview and synopsis of the deliberations at
the meeting, and it describes the outcome.  
 
The major result of the workshop is given next in section 2, followed 
by a brief synopsis of the workshop deliberations in section 3.  
Finally, the milestones for concluding the papers and reporting
the results are given in section 4.
 
Attachment 1 provides the Terms of Reference for the meeting, agreed 
to beforehand by RARE and the FNC.  Attachment 2 gives the working 
agenda, which provided the framework of the workshop.  Attachment 3 
lists the attendees.  Attachment 4 gives an annotated list of documents 
distributed to attendees prior to the workshop.
 
 
 
2. Workshop Conclusions
 
The workshop attendees decided to address the issues of CO/CL 
interworking in a series of 4 documents, the first of which would
serve as an introduction and tutorial.  The following description 
of the approach and the 4 document set is taken from the first document
``An Approach to CO/CL Interworking -- Part I: Introduction'', M.T. 
Rose (editor):

 
          ``Towards a Solution
 
          In the best solution, there is a single mode of OSI network
          service which is truly ubiquitous.  In this case, a single
          community exists and interworking is achieved through the use
          of network-layer relays.  In preparation for this long-term
          scenario, technology must be identified and perhaps
          incrementally advanced to promote a homogeneous network
          service.  In the meantime, a large TCP/IP-based community
          exists and a TP0/CONS community is growing.  Some interworking
          requirements exist today and these requirements are expected
          to increase.  This suggests a short-term solution to address
          immediate requirements, an intermediate-term solution
          applicable as the TP0/CONS community grows large, and a long-
          term solution applicable once two large OSI communities, one
          CO-mode and the other CL-mode, exist and have interworking
          requirements.
 
          Thus, an approach towards the solution consists of three
          parts, and three companion memos have been been written, each
          corresponding to the short-term, intermediate-term, and long-
          term:
 
          In the short-term, one must rely on TS-bridges to provide
          connectivity between non-internetworking communities.  The
          first companion memo, (see [a] below) describes the operation of 
          TS-bridges in such an environment.
 
          The fundamental component of the long-term is the use of
          network-layer relays, supporting either the CO- or CL-mode OSI
          network service.  Use of network-layer relays is well-
          understood.  The second companion memo, (see [b] below), describes 
          how steps can be taken to provide full CONS-interworking.
 
          However, even in the long-term, situations will arise in which
          both network services are required.  In this case, TS-bridges
          are still necessary.  The third companion memo, (see [c] below), 
          describes the operation of TS-bridges in such an environment.''
 
 
 
       [a] M.T. Rose (editor).  An Approach to CO/CL Interworking --
           Part II: The Short-Term -- Conventions for Transport-
           Service Bridges in the absence of Internetworking, CO/CL
           Interworking Workshop, (July, 1990).
 
 
       [b] C. Huitema (editor).  An Approach to CO/CL Interworking
           -- Part III: The Intermediate-Term -- Provision of the
           CONS over TCP and X.25 subnetworks, CO/CL Interworking
           Workshop, (to be published).
 
 
       [c] C. Huitema (editor).  An Approach to CO/CL Interworking:
           -- Part IV: The Long-Term -- Conventions for Network-
           Layer Relays and Transport-Service Bridges in the
           presence of Internetworking, CO/CL Interworking Workshop,
           (to be published).

3. Synopsis of Workshop Deliberations
 
This synopsis follows the final agenda (given in Attachment 2).
 
 
3.1 TUESDAY, JULY 24, 1990
 
INTRODUCTIONS ---
 
Bill Bostwick (Executive Director of the FNC) opened the meeting
with greetings to the attendees,and a discussion of the spirit of
cooperation between U.S. and European networking interests.
 
Following the opening round of introductions, the attendees examined
and agreed on the Terms of Reference and agenda with only minor adjustments.
 
 
PRESENTATIONS ---
 
Note: Hardcopy of all slides and handouts were provided to attendees, and
can be provided by Phill Gross (CNRI) upon request.
 
 --- Kevin Mills (NIST)
 
Kevin Mills described the two main OSI protocol stacks (i.e., CO and CL). 
He also described the current UK and US OSI profiles (ie, GOSIPs) for
utilizing OSI in each respective government; His proposal involves dual 
stack machines on wide-area networks;  He suggests that his proposal may 
look more attractive after we've gone through all the issues involved in 
transport service bridging; He also showed the DoD approach to interworking 
(pure stacks with application layer bridging).  
 
 --- Les Clyne (JNT)
 
Les Clyne described RARE, Eureka, and COSINE.  Briefly, RARE is the association 
of European research networking organizations and their users, with over 20 
member countries in Europe;  Eureka formed COSINE (Cooperation for OSI 
networking in Europe); RARE developed the COSINE specifications).  He showed 
slides of COSINE implementation funding and timetable for pilot projects and 
services.  He also showed IXI service access topology and status.  (IXI 
provides 1984 X.25 services.)
 
 --- Marshall Rose (PSI)
 
Marshall Rose described the transport layer and the CO/CL problem.  He defined 
the Transport-Service Bridge as a solution, which uses transport service 
primitives to ``copy'' from TS-stack to TS-stack.  He gave a good synopsis 
of problems with TS-bridging (see his slides), and concluded by explaining 
terminolgy of ISO DTR "Active transport layer relay"  vs "Passive transport 
layer relay"). 
 
 --- Mitchell Tasman (University of Wisconsin)
 
Mitchell Tasman's presentation was based on the distibuted Wisconsin paper 
(see Attachment 4) which described the Wisconsin implementation of a 
transport-service bridge.

 
 
 --- Christian Huitema (INRIA)
 
Christian Huitema was concerned with the problem of ``locating''the 
TS-bridge.  In current implementations, the originating host is required 
to know which destination addresses need to be reached via a TS-bridge, 
and the originating host also needs to know the address of the TS-bridge.  
He proposed treating transport-service bridge as network level gateway, 
and let IS/IS help solve the problem of locating the bridge.  This removes 
the knowledge and functionality from the end system allows reliance on the 
networking and routing infrastructure to solve for you.  This would not be
a ``near-term'' solution, but it was viewed favorably by many attendees as
a "longer-term" approach for alleviating the necessity of the originating
host to know the TS-bridge address.  He described various methods to cope
with multiple parallel bridges/relays, including NSAP gateway subscription
and modification of TP4.  He also described enhancements to RFC 1006 based 
on the X.25 Subnetwork Convergence Protocol to make TCP look like a special
case of ISO CONS.  The associated routing involved using ISO 10030
(CONS ES Routeing Exchange Protocol).
 
 
DEVELOP LIST OF ISSUES ---
 
Les Clyne led a discussion of issues that needed to be solved.  This 
included both implementation and deployment issues, and short- and long-
term issues.
 
The issues included:
 
        - Determination of TS-stack associated with a network address
        - Ordering of addresses
        - TS-bridge addressing
        - Accounting
        - Access control
        - Load balancing
        - Concatenation of IFUs
        - Availability of address extension facilities (X.25(84))
        - Routeing to the gateway
        - Directory service and location independence
        - Address administration at the gateway
        - Modifications to the transport protocol/use of TP user data
        - Extensions to RFC 1006 to support OSI NS
        - Coordination of 8473 PDU Data Unit identifiers and Transport 
                Reference number allocation
        - Alignment with DTR 10172
        - Impact on end system implementations
 
The requirement to be able to map between the RFC 1006 environment and
TP0/CONS, where addresses on the CONS side are carried in the X.25(84) 
address extension fields, if available, was raised.

3.2 WEDNESDAY, JULY 25, 1990
 
Rose distributed a proposal for separating the TS-bridging problem 
into a short and long term approach.  This proposal was modified in
discussion to encompass a 3 step plan.  In the short term approach, we 
would specify a method for encoding TS-bridge addressing in the NSAP.  In
the long-term, we would focus on Huitema's aproach of relying on the
network layer to locate TS-bridges.  Huitema's other proposal for 
enhancements to RFC 1006 would also be included as part of the transitional
framework to the long-term solution.
 
Other issues, such as security and end-to-end integrity were deemed beyond
the scope of the workshop.
 
The group decided to use Rose's draft ``Implementation Agreements for 
Transport Service Bridges'' (see Attachment 4), as the basis for 
an introduction to the workshop conclusions.  In effect, Rose's paper 
was reworked into 2 documents.  one document would contain a transport 
service bridge tutorial (which would also be applicable for reference 
in the long-term approach) and an explanation of the overall plan.
The specific addressing methodology for the short-term approach would
be placed into a second 2-4 page paper.  
 
Including two other papers to describe 1) the long-term approach, and 
2) the proposed transitional modifications to RFC 1006, the group proposed 
a four document set:
 
      Doc #1 - I) TS-Bridge Tutorial and relation to ISO
               II) Explanation of Strategies for CO/CL Interworking
                  - short-term solution (TS-Bridge Addressing)
                  - Transitional modifications to RFC 1006
                  - long-term 
 
      Doc #2 - Short-term specification 
 
      Doc #3 - Transitional Modifications to RFC 1006
 
      Doc #4 - Long-term Specification
                  a. Subnet Gateway 
                  b. Internet Gateway
 
Rose reworked his draft into documents #1 and #2 for review the following
morning.  Huitema volunteered to draft document #3 at the meeting for review 
on the following morning.  The group agreed on a framework for document #4, 
and Rose produced an outline for the following morning.
 
 
3.3 THURSDAY, JULY 26, 1990
 
On the concluding morning, the group reviewed Documents #1 and #2 by
Rose, and document #3 by Huitema.  Huitema agreed to produced document
#4 after returning to France.  After review and editing by the attendees,
Rose produced preliminary drafts of document #1 and #2, and distributed
these documents to the Workshop mailing list for review.  Finally, we 
developed the plan for reporting the workshop results to CCIRN, RARE, FNC,
and IAB/IETF.  These proposed milestones are given in the next section.

4.  Workshop Milestones
 
 
 
   Documents and workshop report
 
        July 1990       Produce drafts of Documents 1 and 2; mail to COCL
                        electronic mailing list for wider review (DONE)
        Aug  1990       Produce preliminary workshop report for RARE and FNC,
                        Distribute initially to workshop attendees (DONE)
        Aug  1990       Produce drafts of Documents 3 and 4 and distibute
			for review (DONE)
 
 
   RARE/FNC Review
 
        Sept 1990       Present workshop report to RARE CoA and FNC
        Sept 1990       Present workshop results (report and other
			appropriate documents) for review at RARE WG4
        Oct  1990       Present workshop results (report and other
			appropriate documents) for review at CCIRN
        Dec  1990       Present workshop results for approval at RARE 
			CoA and FNC
 
 
   IETF/IAB Review
 
        Sept 1990       Present workshop results (report and other
			appropriate documents) to IETF OSI Integration 
			working group for review 
        Sept 1990       Distribute Documents 1-4 as IETF Internet-Drafts
        Dec  1990       Documents 1-4 reviewed in IETF, recommendation
                        forwarded to IAB
        Jan 1991        IAB considers Documents 1-4 to be published as 
                        Internet RFCs
 
 
 

 
Attachment 1 -- Workshop Terms of Reference
 
 
 
 
 
                        Terms of Reference 
 
                               for
 
                  Workshop on CO/CL Interworking
 
 
                        July 24-26, 1990
 
 
 
The use of different OSI lower layer architectures will act as an
impediment to the development of global OSI connectivity if it is
not possible to interwork between them. This problem is
particularly significant for interworking between the US and
European research communities. Several solutions have been
proposed to provide interworking and there is a need to agree a
specification which meets the service requirements of the
international research community. The CO/CL Workshop has been set up
to recommend a specification that can be used to provide suitable
interworking units.
 
The workshop
 
a)  is asked to develop a technical specification for a general
    purpose TP0/CONS-TP4/CLNS interworking unit, taking into
    account existing or proposed solutions. The requirement to
    interconnect large national or international service networks
    is particularly important.
 
b)  should take into account the ISO work in this area and the
    recommended specification should align with the emerging ISO
    technical report wherever possible.
 
c)  is also requested to advise on the technical specification of
    a convertor which supports TP0/X.25-TP0/TCP interworking
    as there is an immediate requirement in this area to support
    OSI applications.
 
The participants in the workshop will be invited experts from the
US and Europe. It is expected that the workshop will hold one
meeting complemented by the use of electronic mail to support its
activities. The workshop is requested to report to RARE and FNC 
in autumn 1990.
 
 

Attachment 2 -- Workshop Agenda
 
 
                        Agenda and Issues for
 
                    Workshop on CO/CL Interworking
 
                        July 24-26, 1990
 
 
 o Introductions and Administration 
 
        - Introduction by Bill Bostwick (CCIRN Co-Chair, 
                FNC Executive Director)
        - Agreement on chairing and secretarial duties
        - Review Terms of Reference
        - Agreement on agenda and objectives.
 
 
 o Presentations:
 
        - Kevin Mills (NIST) -- US OSI activity
        - Les Clyne (JANET) -- RARE/COSINE OSI activity.
        - Marshall Rose (PSI) -- Transport service bridging.
        - Mitch Tasman (U. of Wiscosin) -- ISO TP4 - TP0 gateway.
        - Christian Huitema (INRIA) -- NSAPs and transparent OSI 
                transport bridges.
        - Others Issues or Approaches?
 
 
 o Determine what protocol stacks need to interwork and what priority the 
        workshop should place on developing solutions for each scenario. 
 
        - Issues -- Is it mainly connection between large CO and CL WAN 
        communities with one or a small number of interworking units in 
        the same geographic location, or do we deal with WAN-WAN 
        connections in different countries, LAN-WAN-WAN-LAN connections 
        with multiple serial interworking units and alternative routes etc.  
        The complexity has an impact on problems of routeing and address 
        management.  On what timescales are the various interworking 
        units needed ?
 
 
 o Develop detailed description of candidate solutions (e.g., Transport 
        bridging, ISO Active Transport Level Relay, etc) for each 
        combination of protocol stacks that need to interwork.  
        Issues to consider include: 
 
 
        - Impact on end systems of proposed solutions.
        - Address management.
        - Routeing considerations.
        - Agreement on further work, assignment of responsibilities, 
                timescales, etc.
 
 
 o Reporting to RARE/FNC.

Attachment 3 -- Workshop Attendees
 
 
 
                    Workshop on CO/CL Interworking
 
                                 at
 
                       Corporation for National 
                          Research Initiatives
 
                           July 24-26, 1990
 
 
 
   Name                 Organization                    Email
 
 
William Bostwick    Federal Net. Council      bostwick@darpa.mil
 
Les Clyne           Joint Network Team        L.Clyne@RUTHERFORD.AC.UK
 
Walid Dabbous       INRIA                     walid.dabbous@MIRSA.INRIA.FR
 
Steve Goldstein     Nat. Science Foundation   goldstein@note.nsf.gov
 
Phill Gross         Corp. Nat'l Res. Init.    pgross@nri.reston.va.us
 
Christian Huitema   INRIA                     Christian.Huitema@MIRSA.INRIA.FR
 
Steve Kille         Univ. College London      S.Kille@CS.UCL.AC.UK
 
Kevin Mills         NIST                      mills@osi.ncsl.nist.gov 
 
Julian Onions       Nottingham University     J.Onions@cs.nott.ac.uk
 
Marshall Rose       Perf. Sys. International  mrose@psi.com
 
Rachid Sijelmassi   NIST                      sijel@osi.ncsl.nist.gov
 
Mitchell Tasman     Univ. of WI - Madison     tasman@cs.wisc.edu
 
 
 
Full postal mailing addresses and phone numbers of each attendee are
available on request from CNRI:
 
	Corporation For National Research Initiative
	1895 Preston White Drive, Suite 100
	Reston Virginia  22132
	Attn: Phill Gross
 
	Phone:  703-620-8990
	FAX:    703-620-5490

 
Attachment 4 -- Distributed Documents
 
There were several documents and position papers distributed to attendees 
that provided a beginning point of view for the workshop.  Many of these
documents were authored by workshop attendees.  An annotation of each 
distributed paper is provided below (based on the description given in
``An Approach to CO/CL Interworking -- Part I: Introduction''.  These
documents can be provided on request by either of the workshop co-chairs.
 
- In "Implementation Agreements for Transport Service Bridges",
  Rose outlined the basic model of transport-layer relaying and
  proposed the basis for an approach in the short-term.  This
  formed the basis for the first document of the workshop ``An
  Approach to CO/CL Interworking -- Part 1: Introduction''.
 
- In "A Survey of Solutions to the Connectionless/Connection-
  mode and the OSI/DoD Interworking Problems", West and
  Sijelmassi outlined the various approaches and assigned
  comparative metrics.
 
- The ISO/IEC Draft Technical Report 10172 (SC 6 N 5906)
  outlined a framework for transport-layer relaying.
  
- In "An ISO TP4-TP0 Gateway", Landweber and Tasman described an
  implementation of a TS-bridge.
 
- In "An NSAP approach to build transport OSI transport bridges,
  Huitema and Dabbous described how ES-transparency can be
  achieved, and in "Extension of OSI TP4 to support transport
  bridging", they described modifications to the TP4 protocol to
  aid in achieving the ES-transparency effect.
 
- In "CO-CL and TCP/IP Interworking and Coexistence", Mills
  suggested that the problem can be solved by using a single
  mode of network service (CLNS), and further that
  interoperability with TCP/IP-based environments would occur
  through the use of application gateways.
 
 
Copies of these documents are available upon request from CNRI:
 
	Corporation For National Research Initiative
	1895 Preston White Drive, Suite 100
	Reston Virginia  22132
	Attn: Phill Gross
 
	Phone:  703-620-8990
	FAX:    703-620-5490