Minutes from SF

Entity MIB (entmib) Minutes
IETF 56 San Francisco
Tuesday, March 18, 2003, 1300-1400
==================================

AGENDA:

Intro and Agenda Bashing -- Margaret Wasserman (5 min)


Document & Milestones Status -- Margaret (20 min)
        - Review outstanding milestones
        - Entity MIB Implementation Report Status
        - Next Steps


Entity State MIB -- Sharon Chisholm (20 min)
        - Open Issues
        - Next Steps


Wrap-up and Action Items -- Margaret (5 min)


Reading List:
http://www.ietf.org/internet-drafts/draft-ietf-entmib-v3-00.txt
http://www.ietf.org/internet-drafts/draft-ietf-entmib-state-00.txt

MINUTES:

Entity Senor MIB Discussion:

No implementations from those in room.  Margaret to poll list.  IF
no implementations available, we will retry in 4-6 months.

Entity MIB Discussion:

Discussion of the current status of the Entity MIB and implementation 
reports.  

CONSENSUS:  Deprecate the LP mapping table and submit for DS.

Discussion of the changes since PS:  new CPU enum and new TC.

CONSENSUS:  Submit the document for DS, even with changes.

Entity State MIB discussion:

Open issues against January 03 version:
	- Dave's email address is wrong
	- Boilerplate needs updating

The document needs to be very explicit that it is a sparse augments in
both directions.  A not-applicable enum would be easier to deal with
than holes, so we will add that.

There are enums introduced that are not in the ITU documents.  Should
give explicit eplanation of what they mean -- perhaps pointer to the
alarm stuff that is their source.

Reference clauses in the TCs to ITU documents.

CONSENSUS:  Agreement to make the three changes listed above.

Editorial comment:  Some terms are used that are not explained first,
for instance reference to a stack (ITU folks might not know what 
this is), locked/unlocked (non ITU folks may not know what this is).
Would be good to add references for terminology.  

Also, there may be subtle differences for terminology about reduncancy 
(primary/secondary and active/backup).

Will try to get an update out by the end of March(?).  Then perhaps
ready for WG last call?

Andy raised issue with possible congestion/redundant traps when an
event happens that affects the whole network.  May have issues when
the a board goes down and all sub-systems also go down (sympathetic
notification issues). 

Concerned about the case of a very large box (10-20,000 interfaces).
Doesn't want to receive 20,000 link downs and 20,000 oper status
downs coming from the box.   

The sympathetic notification issue is actually separate from the
congestion issue.  It might make sense for this MIB to talk about
controlling sympathetic notifications, but the congestion issue
would need to be considered at the overall SNMP level -- maybe
a separate WG?

Issues are related to disman alarm work.  May be able to use
some mechanisms for sympathetic alarm suppression from there.
Also will look at RFC 2863.

Action Items:

Margaret to poll list for Sensor Entity MIB implementations

Andy to update Entity MIB to:
	- deprecate the LP mapping table
	- add conformance with/without CPU enum
	- new boilerplate

Margaret to post implementation reports for review (look
	at the RMON MIB reports)

Sharon and David to update the Entity State MIB to make
	the changes indicated above

Sharon and David to consider possible solutions for the
	congestion/sympathetic alarm problem