CURRENT_MEETING_REPORT_

Reported by Mark Knopper/Merit

Minutes of the TCP/UDP over CLNP-addressed Networks Working Group (TUBA)

Agenda


  o Implementation Status and Demonstration.
  o Document Status.
  o Prioritization of TUBA Work.

     -  Questions asked at Opening Plenary
     -  Dynamic Host Address Assignment
     -  Mobile Hosts
     -  Routing and Addressing Plan
     -  Transition Strategies
     -  Discussion of Technical Advantages of CLNP

  o Demo and Implementation Targets


Implementation Status and Demonstration

The current status of TUBA implementations is:


cisco:        Has telnet and finger initiators and responders, tftp
             initiator, and SNMP agent.  The effort took a long
             weekend, the hardest part being getting the TCP checksum
             right.  Paul Traina indicated that cisco intends to
             modify tftpd to operate over UDP/CLNP as soon as
             operating system support is available.
3Com:         Has telnet initiator and responder.  This work took about
             one week.
BSDi:         Has telnet and SMTP initiators and responders; currently
             a bit buggy.  This implementation is the BSDi
             distribution with Keith Sklower's modified 4.4 BSD
             network code.
NCSA Telnet:     Has telnet and finger initiators; ftp responder works
             for command connection (support for data connection is a
             future work item).
SunOS:        Francis Dupont (at INRIA) has grafted the 4.4 BSD
             modified network code onto SunOS 4.1.2, and has added
             support for UDP over CLNP. No application information was
             available (Francis was not at the TUBA Working Group
             meeting).  Francis has also modified tcpdump to

                                  1





             understand TUBA; contact [email protected] for
             details.
AIX 3.2:      IBM ported the 4.4 BSD modified network code to AIX 3.2.
             Merit will be testing the port.  Yakov Rekhter will
             modify ftp for TUBA after Merit completes the kernel
             work.  It wasn't clear what the status is for other
             applications.


The cisco, 3Com, BSDi, and NCSA Telnet implementations were running in
the IETF terminal room.  CLNP connectivity was available from the
terminal room via an NSFNET EON encapsulator to other TUBA hosts at:


  o cisco via Barrnet
  o 3Com via SURANet and COS
  o NIST via SURANet
  o Merit via the NSFNET
  o LANL via ESNet
  o NORDUNET and other Sites in Europe


Existing Document Status


  o RFC 1347 (the original TUBA proposal):  No identified changes.

  o ``CLNP for TUBA'' Internet-Draft (draft-ietf-tuba-clnp-02.txt):
    Dave Piscitello will polish the pseudoheader checksum calculation
    description.
    Dino Faranacci suggested that we need to think about MTU discovery.
    We might want to use the ER PDU to return the MTU size.
    The idea of padding the CLNP header to obtain word alignment for
    the TCP header was reopened briefly.  It was decided that this had
    already been discussed in the past and we would stick to the
    conclusion that this is not something that can be guaranteed, given
    the number of different subnet services that CLNP operates over.
    Given the implementation experience, the Group decided that it
    would ask for this document to be moved to Proposed Standard.  Dave
    Piscitello will take this as an action.

  o ``Addressing and End Point Identification, For Use with TUBA''
    (draft-ietf-tuba-address-00.ps):  Everyone should go back and
    (re)read this and send comments to the mailing list.

  o ``DNS NSAP RRs'' Internet-Draft (draft-manning-dns-nsap-01.txt):
    This Internet-Draft is the successor to RFC 1348.  It contains a
    better treatment of the inverse mapping for NSAPs than was in 1348,
    but this aspect is still subject to change.  [Note:  Bill Manning
    has posted this Internet-Draft already.]

                                  2





New Documents


  o Catalog of TUBA implementations:  We decided that it would be
    useful to collect the information about what implementations are
    available and who to contact.  Mark agreed to take this as an
    action.

  o CLNP changes from London ISO meeting:  There was a document
    describing possible changes for CLNP that was distributed in a
    recent SC 6 meeting in London.  Mark took the action of getting a
    copy on-line.

  o TUBA Frequently Asked Questions:  In keeping with the theme of
    needing better organization of the TUBA documentation, Mark
    suggested we write a FAQ. Mark will produce a first draft.

  o CLNP Multicast work:  SC6 is working on multicast extensions for
    CLNP and related routing protocols.  Radia Perlman said she will
    ask Dave Oran to post a summary status of this work on the mailing
    list.


Prioritization of TUBA Work

Several questions were asked during the Opening Plenary.


 1. What upper layer changes are necessary?
    The core applications -- including ftp, smtp, telnet, and dns --
    were mentioned.  It was decided that we should create a single
    document that catalogues what changes, if any, need to be made to
    these for TUBA. In most cases, the required changes are minimal.
    Mark agreed to take a first cut at this document.  Dave Piscitello
    agreed to provide the ftp-specific section.  Peter Ford, Yakov
    Rekhter, and Richard Colella agreed to modify ftp from this
    specification.
    Keith Sklower mentioned a draft description of a replacement for
    gethostbyname that he and Eric Allman had devised.  Called
    getconninfo, it is more general than gethostbyname, accommodating
    address families other than AF_INET. This will make TUBA (and other
    IPng proposals) more transparent to the applications.  Keith agreed
    to post the write up as an Internet-Draft.

 2. What is the transition scheme?
    Most of this discussion focused on a problem that John Veizades
    sees:  there is a community of users that do not generally have the
    resources necessary to upgrade their small, older routers to
    accommodate CLNP to support TUBA (e.g., universities).  After some
    discussion it became clear that, whereas some thought that this was
    not a serious issue, John was not convinced.  Dino Faranacci and
    John agreed to take this particular issue off-line.  In any case,

                                  3





    it was clear that the TUBA work needs a transition document to
    answer just this kind of question.  Peter Ford and John Curran
    agreed to draft a transition plan.

 3. Address assignments -- how do we get them?
    This question is fully answered by the NSAP allocation scheme
    outlined in RFC 1237, Guidelines for OSI NSAP Allocation in the
    Internet, July, 1991.  There is already a well-defined method of
    obtaining and assigning NSAP addresses.  In the U.S., address space
    can be obtained from either the US GSA or from ANSI.

 4. How does TUBA address mobile hosts?
    Deferred due to lack of time.

 5. Are there any known boundary conditions?
    There were no known boundary conditions involving TUBA.

 6. What about Scaling?
    In response, reference was made to a seminal paper from 1971 by
    Kleinrock and ?.
    Stev Knowles asked, ``What if you have one million networks?  How
    does CLNP and it's routing protocols handle this?''  A lively
    discussion ensued; there was not a specific response as it's a
    complex question.


It was agreed that the TUBA Working Group should discuss the topics of
scaling and mobile hosts.

Discussion of Technical Advantages of CLNP

Radia Perlman wanted to make the point that we need to recognize the
technical strengths of CLNP. She enumerated three in particular.


 1. Autoconfiguration -- By using a unique System ID in the NSAP, it is
    relatively easy to do address autoconfiguration.  This would
    greatly reduce administrative overhead in assigning and changing
    addresses, and allow for easier portability of systems.

 2. Infinite scaling property -- Given the size and flexibility of NSAP
    addresses, address prefix routing provides a large number of
    potential levels in the routing hierarchy, assuming that prefixes
    are based on nibble boundaries.

 3. ``Free'' routing across WANs -- Embedded subnet addressing can be
    used to simplify routing in environments that make use of WANs for
    interconnection.  This entails assigning NSAPs with a WAN-based
    subnet address in the high-order part of the NSAP. The WAN-based
    part of the subnet address would then be used to perform the
    cross-WAN routing hop (e.g., from one routing domain to another,

                                  4





    both connected to the same WAN). Note that domains not connected to
    the same WAN would continue to route using the normal routing
    protocols (i.e., ISIS and IDRP).


Dynamic Host Address Assignment

One part of the solution to dynamic host address assignment is ES-IS,
which is reasonably straightforward.  Bill Warner agreed to draft text
that describes how ES-IS is used to do dynamic address assignment.

Another part of dynamic host address assignment is how to get the
information into DNS. This is not so obvious.  John Curran agreed to
write some text for this.

Routing and Addressing Plan

Ross Callon wrote a routing and addressing Internet-Draft for TUBA in
October.  Everyone was assigned to (re)read this and comment (see
Internet-Draft draft-ietf-tuba-address-00.[txt,ps]).

The subject of globally unique EIDs was raised once more.  There was
violent agreement that we should do this in the NSAP System ID field.
However, there was some disagreement on the mechanics.  Ross suggested
mandating that the System ID field be taken from a single,
globally-coordinated 48-bit number space (*not* synonymous with IEEE MAC
addresses).  Keith had a somewhat different idea, allowing variable size
EIDs and, hence, variable sized System IDs.  Each proponent was asked to
write a short description of their proposal and post it to the mailing
list.  Dave Piscitello agreed to write up Ross's proposal.

Demonstration and Implementation Targets

It was recognized that the TUBA demonstrations could benefit from better
planning and coordination.  George Chang agreed to take the lead in this
area.

Summary of Action Items


Dave Piscitello     CLNP for TUBA document (update) and submit for
                   Proposed Standard.
                   FTP for alternative network layers:  Specification.
                   The implementation will be done by Peter Ford, Yakov
                   Rekhter and Richard Colella.
                   EID administration text.
Ross Callon         Addressing doc (update), comments needed from Group.
Manning and Colella   DNS for NSAPs I-D (RFC 1348 update).

                                  5





Mark Knopper        Catalog of TUBA implementations.
                   CLNP changes from London ISO meeting (make document
                   available).
                   TUBA Frequently Asked Questions.
                   Application changes document (what needs to change
                   for each).
                   ``Cron job'' to update the Group on status weekly.
                   (This item refers to the offer Mark made to remind
                   the Group periodically on the status of each action
                   item and what is left to be done.)
Radia Perlman       Status of CLNP Multicast work.
Paul Traina         Tftpd - implementation.
Keith Sklower       getconninfo Internet-Draft (replacement for
                   gethostbyname).
Ford and Curran     Transition document.
Bill Warner         Autoconfig (dynamic host address assignment using
                   ES-IS), specification.
John Curran         NSAP insertion into DNS text.  The implementation
                   will be handled by Dave Piscitello.
George Chang        Demo PR and coordination.



Attendees

Philip Almquist          [email protected]
Jim Barnes               [email protected]
Russell Blaesing         [email protected]
Rebecca Bostwick         [email protected]
George Chang             [email protected]
John Chang               [email protected]
Enke Chen                [email protected]
William Chimiak          [email protected]
Richard Colella          [email protected]
Michael Collins          [email protected]
John Curran              [email protected]
Dino Farinacci           [email protected]
Eric Fleischman          [email protected]
Francois Fluckiger       [email protected]
Peter Ford               [email protected]
Vince Fuller             [email protected]
Peter Furniss            [email protected]
John Gawf                [email protected]
Eugene Geer              [email protected]
Tony Hain                [email protected]

                                  6





Susan Hares              [email protected]
Woody Huang              [email protected]
David Jacobson           [email protected]
Laurent Joncheray        [email protected]
Mark Knopper             [email protected]
Paul Lustgraaf           [email protected]
Carl Madison             [email protected]
Tracy Mallory            [email protected]
Bill Manning             [email protected]
Jun Matsukata            [email protected]
David Meyer              [email protected]
Dennis Morris            [email protected]
Matthew Morrisey         [email protected]
Peder Norgaard           [email protected]
Laura Pate               [email protected]
Maryann Perez            [email protected]
Radia Perlman            [email protected]
David Piscitello         [email protected]
Willi Porten             [email protected]
Yakov Rekhter            [email protected]
Ben Robinson             [email protected]
Yzhak Ronen              [email protected]
Michael Safly            [email protected]
Paul Serice              [email protected]
Roxanne Streeter         [email protected]
Steve Suzuki             [email protected]
Wayne Tackabury          [email protected]
John Tavs                [email protected]
Kamlesh Tewani           [email protected]
Richard Thomas           [email protected]
Paul Traina              [email protected]
John Veizades            [email protected]
William Warner           [email protected]
Linda Winkler            [email protected]
Cathy Wittbrodt          [email protected]
Charles Young            [email protected]



                                  7

You are viewing proxied material from ftp.icm.edu.pl. The copyright of proxied material belongs to its original authors. Any comments or complaints in relation to proxied material should be directed to the original authors of the content concerned. Please see the disclaimer for more details.