CURRENT_MEETING_REPORT_ Reported by Dean Throop/Data General Minutes X.25 Management Information Base (X25MIB) The X25mib Working Group met at the IETF meet in Santa FE on Monday, November 18th, 1991. All draft documents were discussed and all were referred back to the editor for further changes. Inter-MIB Structural Issues Fred Baker raised the issue of a single X.25 packet layer running over multiple LAPB sessions. The current X.25 MIB will not support such a structure. Since the one over several structure occurs in other situations, it was agreed a general solution would be better than putting direct support for such structures into the X.25 MIB. Fred Baker volunteered to draft a MIB that would allow one level_X_MIB to be redirected to identify multiple level_X-1_MIBs. To facilitate this, the SYNTAX of the x25InfoDataLinkId will be changed from INTEGER to OBJECT IDENTIFIER. That object identifier will identify an instance of the index for the first table of the MIB for the layer under that X.25. For X.25 running over LAPB, it will be an instance of lapbParamIndex. For X.25 running over multiple link layer entities, it will be an instance in the table of the MIB that Fred will draft (see above). For X.25 running over interfaces that don't have specific MIBs, it could also be the ifIndex for an interface. A similar change will be made in how the HDLC MIB identifies the port below it. The hdlcParamPortIndex will be deleted and the SYNTAX of hdlcParamPortId will be changed to OBJECT IDENTIFIER. The object identifier will identify an instance of the index for the first table of the MIB for the port under LAPB. In general this will be an instance of rs232PortIndex. Dave Perkins said he had a new tool which identified several syntax problems with the current drafts. It was agree the MIBs should be changed to correct these problems and the issue was referred to the editor to complete. HDLC MIB It was agreed to change the name of HDLC back to LAPB because the MIB wasn't broad enough to cover all variants of HDLC. It didn't cover basic HDLC framing, nor SDLC, nor LAPD. It is indeed a LAPB MIB and should be so named. The LAPB MIB will be expanded to include support for ISO 8885 XID negotiations. Some of the attendees with European experience indicated that XID negotiations are important for that community. Conformance 1 The issue of conformance was discussed. It was agreed that the MIBs will contain tables that are mandatory and optional. A vendor must implement all mandatory tables to claim conformance. The optional tables will be present to allow management of implementations that implement more than a minimal X.25/LAPB stack. X.25 MIB With the conformance issue in mind, the X.25 MIB will be examined to restructure the tables to make some tables optional. The objects required by a minimal X.25 implementation should be in required tables and all other objects should be in optional tables. The X.25 MIB will also be examined to determine if some objects can be eliminated. Herve Goguely from LIR Corporation volunteered to review the current MIB in light of his European experience and develop a list of objects to consider deleting. The group discussed recording error conditions from the last closed connection. It was agreed a table should be added to record the reason for the last abnormal close. The table should allow recording the last N conditions however vendors will only be required to keep 1 condition; vendors may choose to keep more if resources permit. The RMON MIB will be examined for a possible paradigm for structuring the table. IP over X.25 Andrew Malis informed the Working Group that the IPLPDN Working Group has started writing a new RFC to replace RFC 877. He said there were several aspects of that draft that were inconsistent with the IP over X.25 MIB. The IP over X.25 MIB will be examined to align it with the revised RFC on IP over X.25 coming from the IPLPDN Working Group. Andrew Malis and Fred Baker will serve as liaison between the IPLPDN and X25MIB Working Groups. Attendees Fred Baker fbaker@emerald.acc.com James Davin jrd@ptt.lcs.mit.edu Herve Goguely lircorp1!roger@uunet.uu.net Cheryl Krupczak cheryl@cc.gatech.edu Andrew Malis malis@bbn.com Evan McGinnis bem@3com.com David Perkins dperkins@synoptics.com Philippe Roger lircorp1!roger@uunet.uu.net Bob Stewart rlstewart@eng.xyplex.com Mark Therieau markt@python.eng.microcom.com Dean Throop throop@dg-rtp.dg.com 2