[Nml-wg] Meeting summary

[Freek Dijkstra]

Here is a very brief summary of my notes made during the last meeting.
Martin made some comments to. If anyone else has additions or
corrections, please sent it to the list.


Regards,
Freek Dijkstra



AGENDA / TOPICS DISCUSSED
=========================
- deliverable 2.
- discussion if we should define a protocol to exchange topologies.
- discussion on adding time-based information to the schema
- discussion on multi-layer extensions to the schema

- presentation by Freek on G.805 and NDL multilayer
- presentation by Martin on UNIS (aka)
Slides of both presentations are available at:
http://forge.gridforum.org/sf/docman/do/listDocuments/projects.nml-wg/docman.root.meeting_materials.ogf26_chapel_hill

- discussion on services
- some additional changes to the schema


DELIVERABLE 1
=============

We did not discuss this.

I once again ask one of the authors to step forward to make the two
minor wording changes, and submit it (please email to the list, I just
noticed the document is locked on gridforge).

Otherwise, I'll make the changes, but that will be in july, as june is a
busy month for me.


DELIVERABLE 2
=============

We decided to have conference calls between the authors.
We will use the MCU (H323?) video conferencing facilities (available to
Nordunet and SURFnet amongst others) with one of the existing audio
bridges so people can also phone in.

Jeroen will propose a time and date. I propose this is announced on the
list.

If there are volunteers who can figure out the details how to set up
such audio conference, please step forward.

USE CASES
=========

I (Freek) received some additional use cases:

- backplane limitation use case from Chin Guak + Ralph Niederberger
- Received use case document from Jeff Boote
- John Vollbrecht expressed interest in path finding usage (this is
mostly a usage area, not a practical use case, but it may be good to
turn this in one or more practical use cases).

I am interested if people can adopt a use case and check if that
particular use case is supported by the schema.

If I (as author of that section) were to do each check, it is likely
that I miss something, as I can't look into the minds of whoever made up
the use case, and may have a different interpretation.

PROTOCOL DEFINITION
===================

Topology description and abstraction are definitely in scope of the NML-WG.

The NSI WG asked both the NML-WG and the NMC-WG groups (very informally)
if it is in their scope to define a topology exchange protocol, next to
the schema.

Some remark, and I am certain that I will misquote some people here:

Jeff: not in scope for NML.
Evangelos: any current topology implementation is not much work.
Nevertheless,
Evangelos: in scope of NML, NSI and NMC.
Guy: worried that we do things twice (in multiple working groups).
John: many ways in which topology exchange can be done. Not trivial.
Separate work item.
Jeff: were we do this work depends on the overlap of people who are
interested.

There was no clear consensus. Topology exchange (method and protocol)
may in scope of the NML-WG. We will see how things develop. If it is
trivial, we do it in the NML. Complexity may be caused by complex
queries involving other resources (end-hosts, data, ...). If that is an
issue, the effort should be done in a new WG. Jeff argues that this is a
portion of NMC.

CHANGES TO THE SCHEMA
=====================

About 6-8 proposals to change the NML schema have been (very shortly)
discussed. See separate email.

SERVICE DEFINITION
==================

We discussed this for about one-and-a-half hour. I'll try to summarize
that in a separate email later this week.