[Nml-wg] Minutes conf call

[Freek Dijkstra]

Here are my minutes. I'm sure some of my interpretation is wrong, so
corrections are welcome.

NML conference call
Feb 10, 2011.

Participants:
Aaron Brown
Freek Dijkstra
Chin Guok
Guy Roberts

Excused:
Roman Łapacz
Jason Zurawski

Status Circuit Monitoring Architecture
======================================
Roman, Aaron and others are discussing (off-list) an architecture for
circuit monitoring (this is an improvement of the current E2Emon
architecture). Aaron gave some background on it's status and how it
relates to NML.

There is mostly agreement in the discussion. Schema fragments are mailed
back and forth.

The architecture considers a topology with a circuit, which is composed
of segments (each segment is in a different domain). The idea is that a
client uses the circuit identifier to query information in a topology
service, and receives (among others) a list of segment identifiers. It
then uses a lookup service to know which domain to query for each
segment, and can then ask detailed information at each domain about each
segment.

The current discussion is on the protocol and details, but the topology
requirements are mostly clear. Aaron will present this as a use case to
the NML-WG. It seems that we're mostly in agreement here already.

Status NSI-WG
=============
Guy and Chin gave some background on the status of the NSI working group.

The NSI working group is currently discussion the STP and SDP definition.
An STP is a demarcation point between two domains. One of the questions
is how this maps to the topology in NML.

The NSI-WG thoughts are in framework document (GFD.173). Jeroen van der
Ham went through it and gave useful feedback.

The STP is traditionally seen as (a pointer to) a NML port.
In recent discussion, it is wondered if a link can also be a demarcation
point, thus if a STP can also (point to a) fiber. This may be the case
for trans-oceanic fibers that a managed by a third party.

Discussion is not finished. It is probably not required to get involved
at this stage. As soon as the NSI-WG has reached consensus, they will
present that as a requirement to the NML-WG.

Status serial compound call #1
==============================
Freek requested input on his email of February 2 for those who had not
replied.

There was no request for clarification at the call, and Freek indicated
that he would be happy to provide details later if required.