Routing Information Protocol Working
Group (RIP)-Gary Malkin, Chair
The RIP Working Group met on Tuesday, 3/5,
from 1730-1800. We began with a presentation of the agenda and
of the current documents associated with the group's charter.
RIP-2 (RFC 1723) is still waiting for the
inclusion of introductory material from RFC 1058. Once this is
done, RIP-2 (all RFCs) can be submitted for consideration as full
The RIPng Internet Draft has been updated
with comments from the last call. A few more things need to be
added before a new Internet-Draft can be published. Following
a new last call, it will be ready for submission for consideration
as a Proposed Standard.
Gerry Meyer will make a few editorial changes
to the Triggered RIP Internet-Draft. After the new Internet-Draft
is published, it will be submitted for consideration as a Proposed
Open Shortest Path First IGP Working
Group (OSPF) -John Moy, Chair
Rob Coltun lead the meeting of the OSPF
Working Group that met on Monday and discussed the Opaque LSA,
Cut-Through Advertisements and OSPF Version 3 For IP Version 6.
Rob Coltun presented the following two proposals.
The updates to the current draft were given which included:
A brief overview was given of the Cut-Thru
Advertisement draft (which is soon to be released).
No adverse reactions to the above two proposals.
Rob and Dennis then presented the OSPF version
3 draft. Besides for the issues of cleaning up the draft the only
issue was to put the tag back into the AS external LSA which is
handy for policy filtering. The view of the room was that the
protocol was very well designed and that the implementation of
version 3 from version 2 would probably be smooth.
After the meeting a few vendors said that
they would start implementations fairly soon and were interested
in feeding comments into the draft in the short term.
Two important issues were brought up after
the meeting which were:
Inter-Domain Routing Working Group -Sue
Hares and Yakov Rekhter, Chairs
Working group met with agenda:
Status of documents:
BGP-4 MIB (draft-ietf-bgp-mibv4-06.txt)
awaits experience report to IESG to be forward to Proposed. The
document IDRP for IPv (draft-ietf-idr-idrp-v4v6-02.txt) will
be merged with the IDRP specification. The documents related to
Destination Preference Attribute (DPA) (draft-ietf-idr-bgp-dpa-05.txt,
were discussed. draft-ietf-idr-bgp-dpa-05 will be revised to reflect
working group consensus, and will be submitted for proposed standard.
the other two documetns will be submitted for informational status.
The BGP community specification (draft-chandra-bgp-communities-00.txt)
has been release for over a year. It has been implemented by Bay
networks and Cisco. It is suggested that Yakov make the last call,
and move it to proposed standard.
BGP Confederations document (draft-traina-bgp-confed-00.txt)
is moved to experimental. Route Reflection specification (draft-bates-route-reflect-00.txt)
recommends that this move forward to experimental.
Routing Over Large Clouds Working Group-Andrew
The ROLC Working Group met in two sessions
at this IETF. There were 142 attendees.
The working group received an ATM Forum
MPOA report. The 08-beta version of the NHRP specification was
discussed, and some open issues were resolved. The chair announced
that the ipatm and rolc working groups will produce one common
multiple server synchronization mechanism for use by NHRP, ATMARP,
and MARS, specified in a stand-alone document. Keith McCloghrie
presented the ATM Forum's LAN Emulation Server Synchronization
protocol. Jim Luciani presented his Server Cache Synchronization
Protocol (SCSP). He and Joel Halpern are collaborating to merge
SCSP with the synchronization protocol in the ipatm Classic2 document.
Maria Greene and Jim Luciani reported that
they have received from the MIB from Avri Doria, the previous
editor. It is based in part on the Classical IP MIB. Derya Cansever
reported on the status of the NHRP Protocol Applicability Statement.
Yakov Rekhter reported on the status of his NHRP for Destinations
off the NBMA Network draft. Yakov presented a mechanism, Sparse
Mode PIM over ATM, to eliminate extra layer three hops across
an ATM network for multicast traffic. Masataka Ohta discussed
the inscalability of multicast over large clouds. Rob Coltun presented
a mechanism for using information in OSPF to eliminate some hops
for NHRP requests. Following Rob's talk, the WG adjourned.
Nimrod Working Group-Noel Chiappa, Isidro Castineyra and Dave Bridgham, Chairs
The Nimrod working group met on Wednesday
March 6 from 7pm to 9pm. The bulk of the time was spent going
over the current state of the architecture and protocols. There
was also a report on the state of the implementation at BBN.
Inter-Domain Multicast Routing -
Anthony Ballardie and Bill Fenner chairs IDMR met in two sessions. The first session had Tom Pusateri talking about the DVMRP spec, in which he specified DVMRP version 3, which mrouted uses, except he removed all of the mistakes that the mrouted implementors made from being part of the protocol. Bill Fenner talked about IGMPv2, the most significant change being the requirement of the IP Router Alert option. Deborah Estrin talked about PIM and recent changes, mostly simplifications in the elimination of RP timers and the introduction of an algorithmic mapping from group address to RP. Dave Thaler talked about the characteristics that such a mapping should have and presented the current implementation. Deborah then presented the PIM<>DVMRP interop architecture. The second session had Tony talking about CBT updates and CBT interop, which elects a single designated border router to inject traffic into and out of a CBT domain. The transit issue is not addressed. Scott Reeve presented the CBT MIB, which several people thought had several items duplicated from other MIB's. Bill Fenner presented new features in DVMRPv4, including source-less prunes, source-specific prunes, prefixes in routing updates and a prune-is- expiring message. Yuan-Chi Chang presented a new proposal called "Core-Group Based Trees", which dynamically locates cores based upon group membership, and in which the top-level cores unicast traffic to all other cores. Shridhar Shukla presented an architecture for hierarchy that groups nodes together and builds multicast trees from the nearest border router to the next level, and requires registration of multicast groups in databases at each level.