IETF Operations Area Open Meeting and Operations Area
Working Group Combined Meeting

6 November 2012, 1300-1500, Atlanta, GA

Minutes by Joel Jaeggli, edited by Melinda Shore

1) Agenda bashing, administrivia

2) opsarea status update

3) IETF/IEEE 802 cooperation, Dan Romascanu

   Dan provided an overview and history of cooperation with
   IEEE 802, including Ethernet MIBs, IPFIX, and Radius.
   Benoit asked participants to let him know if there are
   places needing additional cooperation

4) draft-ietf-eman-rfc4133bis, Dan Romascanu

   Status update on revision of ENTITY-MIB-V4, summarized
   major changes.  Primary point of contention was
   concerning whether or not there should be a procedure to
   deprecate entries in IANAentPhysicalClass.  Sense of the
   room was there should not be.  [n.b. since the meeting it
   has been determined that this was raised as the result of
   a misunderstanding, and is not an issue]

   Requested that the IESG publish the draft
   as a standards track RFC.

5) opsawg status update
  
   draft-ietf-opsawg-firewalls is still in development, new
   draft has major open sections

   draft-ietf-opsawg-lsn-deployment authors would like the
   draft to go into WG last call, agreed that there should
   be another round of discussions and feedback first.

   draft-ietf-opsawg-automated-network-configuration draft
   has received a lot of DISCUSSes and will need major
   revision.

6) draft-ietf-opsawg-firewalls, Fred Baker

   Overview of the document structure and sections.  Scott
   Bradner asked for the elevator pitch - what is this
   document for?  Fred mentioned that it's come up in v6ops
   and homenet.  If every time the word "firewall" comes up
   we have to have this discussion we should have consensus
   as to what it is.

   Phillip Mathews asked if the intent was to standardize
   firewall behaviors, comparing this effort to the behave
   working group.  This is different from behave because
   behave was trying to get STUN to work across a variety of
   NATs with widely varying behaviors.  We are not trying to
   solve that problem with regard to firewalls.

   Curtis Villamizar asked that the document included some
   text stating that firewalls are ineffective, but that
   suggestion fell to the floor with a loud "thud."  Victor
   Kuarsingh asked for some clarification as to what is
   considered a firewall and firewalling, since a variety of
   devices in a network do some sort of policy-based
   filtering. 

7) draft-ietf-opsawg-lsn-deployment, Victor Kuarsingh

   Authors feel draft is mature and has received sufficient
   feedback for working group last call.  Melinda said she
   wasn't conformtable with the level of review.  Three
   people volunteered to review the document: Tina Tsou,
   Chris Donnelly, and Chris Liljenstolpe.  If the reviewers
   are happy we'll send it to WG last call.

8) draft-krishnan-opsawg-large-flow-load-balancing, Ram Krishnan

   Identified problems with distribution of flows across
   load balancers, proposed a "best practice" document
   describing technique for doing better.  There was
   skepticism in the room about whether or not their
   proposal would actually work, but Scott Bradner
   identified the "high-order bit" as the work being
   interesting but needing feedback from operators.  Chris
   Liljenstolpe said it would be helpful if vendors could
   get their customers to participate in opsawg and provide
   feedback.

9) draft-ersue-constrained-mgmt, Mehmet Ersue

   Overview of rationale for the work, discussion of changes
   since IETF 84, identification of additional areas needing
   text.  Contributions are welcome.  Mailing list
   discussion at https://www.ietf.org/mailman/listinfo/coman 

10) draft-shao-capwap-plus-ps and draft-cao-capwap-eap, 
      Cao Zhen

   Background on CAPWAP and description of interoperability
   problems between access points and access controllers
   from different vendors, proposes EAP encapsulation in
   CAPWAP control plane.  Asked that CAPWAP work be
   restarted in opsawg, specifically: 1) EAP encapsulation,
   2) air interface management extension, and 3) BCP on
   local/split MAC (the latter may be an individual
   contribution).  Discussion of whether or not this may be
   an IEEE liaison issue.  Conclusion was that it's not,
   since CAPWAP is an IETF product and doesn't need a
   liaison from IEEE.  Benoit encouraged a new draft with
   two distinct problems.  We'd need to recharter to work on
   those, either restarting capwap or adding it to the
   opsawg charter.

11) Nothing raised at opsarea open mike.

Many thanks to Joel Jaeggli for his notes.