minutes of San Jose meeting RIPv2 WG

[Gary Scott Malkin <[e-mail]>]

CURRENT_MEETING_REPORT_

Reported by Gary Malkin/Xylogics, Inc.

Minutes of the RIP Version II Working Group (RIPV2)


Status Update


Chairpersons		Gary Scott Malkin / [e-mail]

Mailing List		ietf-rip(-request)@xylogics.com
Archives		xylogics.com:gmalkin/rip/rip-arc

Date of meeting		San Jose IETF / December 6, 1994

Progress		Received general approval of "RIP-II Cryptographic
			Authentication" specification.  Agreed to proceed
			with development of RIPng for IPv6.  RIP-2 and
			Demand Circuit RIP awaiting timeout for advancement
			in the standards track.

Agenda

	1 - Review New Charter
	2 - RIP-2 Status
	3 - Review of "RIP-II Cryptographic Authentication" Internet Draft
	    Open Issues: backwards compatibility
	4 - Discuss RIPng
	    Should there be a RIPng?  Perhaps a DVng?  The RIPng Internet
	    Draft may also be discussed.
	5 - Any other issues
	6 - Summary of decisions and action items

Yesterday, the Router Requirements WG decided that RIP-1 should be moved to
Historic status, now that RIP-2 is a Draft Standard.  Joel Halpern, the
Routing Area AD, requested that the RIP WG make the same request.  There
were no objections to the motion, so a formal request for the status change
will be issued by this WG.

Discussion of the new charter was put off pending a decision on the
desirability of implementing RIPng.

The first major topic discussed was the desirability of implementing RIPng.
This discussion was prompted by the belief that RIP served only to allow
hosts to learn about routers, a feature relatively new to IPv4 but an
intrinsic part of IPv6.  Most people attending the meeting, however, believe
that RIP serves a useful niche as a routing protocol, despite the existance
of OSPF.  The argument is that OSPF requires far more effort to implement
and configure than RIP, and that OSPF is far more CPU and memory intensive
than RIP.  Further, there was the general opinion that if RIPng is not
created as an IETF standard, most vendors will implement a RIP-like protocol
which may not interoperate with other vendors' RIP-like protocols.  This
is the situation which prompted the creation of RFC 1058.  It was generally
agreed that such a situation should not be allowed to occur in IPv6.

Ran Atkinson gave a presentation on the "RIP-II Cryptographic Authentication"
Internet Draft.  It was generally viewed as a "good thing."  In a few weeks,
to allow time for any comments/corrections, the draft will be submitted to
the IESG for consideration as a Proposed Standard.

The final topic of discussion was a review of the "RIP for IPv6" Internet
Draft.  This was a Drafty Draft, as there was some doubt as to whether or not
there would be a need for any such protocol.  The major comments were:
	- remove the throughput class
	- indicate that RIPng packets may be carried across IPv4 and IPv6
	- indicate that RIPng/IPv4 should support authentication
	- indicate that RIPng/IPv6 does not need authentication because
	  such is incorporated into IPv6
Additionally, Keith Sklower proposed a mechanism to reduce the size of some
RIPng entries by making use of the fact that the higher half (or so) of the
destination route addresses will usually be identical.  A lot of discussion
regarding RIPng will need to occur on the mailing list before the next draft
is ready for the Danvers meeting.