Minutes of ADSLMIB - 53rd IETF Minneapolis
20 March 2002
Minutes recorded by Mike Sneed.

The ADSLMIB Working Group met at 1:00pm on 20 March 2002.  The meeting followed the posted agenda.

The meeting began with announcement that the mailing list has been down for about a month.  Inovia is no longer able to host the list, and arrangements have been made to have the IETF secretariat take over administration of the list. The last backup of the list itself is a couple of months old, so not all subscribers may be notifiable when the new list is up.

Document status was reviewed.  The ADSL Extension MIB is awaiting a final draft.  The SHDSL MIB is in the RFC Editor queue.

The appropriate path for the 64 bit Performance History TC document (once known as RFC2493bis), which is needed for the VDSL MIB, was discussed. The possibility of updating RFC2493 with a set of 64 bit counters was rejected because documents which reference it are attempting to advance in the standards track. The consensus was that a second standards track document containing the TCs should be produced within ADSLMIB. It was pointed out that the semantics of the performance counter are not consistent with Counter64.  It was noted that this is part of a larger problem with 64 bit objects, and that therefore the document will be flawed.  The consensus remained that ADSLMIB should produce a 64 bit TC document for its own use, and which could be used by others who are not deterred by its flaws.

Bob Ray led a discussion of the current VDSL MIB work.  Rajesh Abbi will join him as a co-editor of this document. The DSL Forum is currently working on a protocol independentlist of management objects for VDSL.  Concern was expressed that there has not been much feedback on the mailing list, and that this may indicate that there is little interest in a standard VDSL MIB or that the DSL Forum and IETF may be proceeding independently. The chair will "liaison" letter to the DSL Forum to solicit input on the list.

It was again asked what percentage of DSL equipment is being managed by SNMP vs CLI and other mechanisms.  The consensus was that SNMP is widely used, along with TL1, particularly in flow-through environments.