Ops Area Working group Thursday 11 November 2010

Minutes by Fred Baker

1 - welcome, note well, scribes, status

2 - IANA Reserved IPv4 Prefix for IPv6 Transition        Jason Weil
    draft-weil-opsawg-provider-address-space

Chris presented the question before the house: shall the IETF ask IANA
to allocate a prefix for operators to use in the NAT444 domain. The
proposal is controversial. The question is to be called in v6ops Friday.

Outcomes:

The draft needs to at least reference previous drafts if not report on
their discussions. Draft needs more context and description

Some proposed having ISPs squat on routable but unrouted Internet
address space; the sense of the room is that the IETF really should not
recommend that.

Straw poll: more people in the room supported the draft than didn't,
perhaps 2:1. About half the room expressed an opinion.

3 - Dynamic Virtual Network Engine                Margaret Wasserman
    draft-mrw-dvne-fw
    draft-mrw-dvne-prot

Issue being addressed in dynamic virtual networks - security
configuration, firewalls, NATs, and other middleware, and in some cases
routing with BCP 38, and interrupt the ability for any-to-any
communications. Management can also be difficult.

Need a service provisioning protocol. Comment from Dan Romanescue - what
is a service? What is provisioning? What requirements exist for a
protocol?

Other SDOs are important: BMF, TMWG.

Discussion taken to the list

4 - OAM Acronym Soup                                  Scott Mansfield
    draft-ietf-opsawg-mpls-tp-oam-def

Brief discussion, uncontroversial, headed for working group last call on
the list.

5 - An Overview of Operations, Administration, and Maintenance (OAM)
    Mechanisms                                 Yaacov Weingarten
    draft-ietf-opsawg-oam-overview

Merger of two pre-existing drafts, and is intended to provide simple
connectivity verification and perhaps route trace in a network that
combines IPv4/IPv6 with MPLS.

BFD is a little different - verified bidirectional continuity check.
Since it requires an ongoing session, while it could be used, it seems
heavy-weight.

VCCV is a TDR for a network - fault detection and diagnostics for
network paths.

IEEE 802.3ah - First Mile Ethernet can be used to monitor a link.

IPPM has tools for such things as well

ITU-T Y.1711 (MPLS OAM) is specific to MPLS.

Outcome requested: can't be published due to dependence on ongoing
MPLS-TP work. Document to be recycled to avoid expiration, but expect to
send to IESG in April.


6 - Problem Statement for the Configuration of Large-Scale IP Networks
    draft-tsou-opsawg-network-configuration            Tina TSOU

Problem statement expected to be completed in December, solution likely
to be done in March. General sense is to adopt as a working group draft.

7 - An Overview of the IETF Network Management Framework and Standards
    draft-ersue-opsawg-management                   Mehmet Ersue

Draft being developed for NIST SGIP in response to CSWG and SGIP
Governing Board request, and presumably useful for other purposes.
General sense is supportive. The suggested result is something that
outlines various technologies and identifies their target purposes and
pros/cons.

8 - A method to monitor MPLS label mappings        Gregory Cauchie
    draft-cauchie-opsawg-monitoring-mpls-label-mapping

Discussion resulted in no actionable outcomes, but the author is
encouraged to continue. Generating new mechanisms for reporting data is
not what the operators need, but gathering the information in ways
similar to that suggested has value.

9 - Internet facing server logging recomendations       Alain Durand

NAT444 CGN to be deployed widely soon. Logging recommendations
recommended to be BCP and picked up in servers for diagnostic purposes.

Alain wants a home for this somewhere in the IETF. It is complementary
to work in behave, so behave is a possible option. However, absent that,
support was demonstrated for accepting it here.

10 - Information Elements for Flow Performance Measurement
    draft-akhter-ipfix-perfmon                          Aamer Akhter

Various internet drafts and RFCs exist to support flow measurement for
SLA support - fault diagnosis and early warning. Not directly about
specific applications (RTP etc), but uses RTP as a way to perform
measurement.

Looking for a home. The chairs asked if there was real interest in such
a technology; several commented that there was real operational usage of
similar proprietary technologies, and that operators find value in such
a measurement.

The chair's advice was that the work is interesting and appears useful,
but needs some work before being adopted.

Open Mike time:

Dan Romanescue (AD) said he really enjoyed the meeting. He wondered
why...