Re: GLIF meeting presentation

[Ronald van der Pol <Ronald.vanderPol@xxxxxxxx>]

On Thu, Aug 21, 2008 at 23:38:27 -0400, Tom Lehman wrote:

> Ronald,
> 
> I am looking at the overall GLIF agenda:
> http://www.glif.is/meetings/2008/
> 
> Looks like there are simultaneous sessions for the Wed 15:30-17:00 and Thurs
> 09:00-10:30 time slots.
> 
> Who do we expect to be at the Global Identifier Task Force meeting on Thurs
> 09:00-10:30? I guess it is just the people who are specifically interested
> in this topic?  I am guessing the GNI Specifications Task Force is going to
> take alot of people away.
> 
> I know I will need to go to a significant part of the GNI Specifications
> Task Force, since that has much to do with the control plane activities I am
> involved. That does run at same time as Global Identifier Task Force.

I hate these parallel sessions. I also want to follow the GNI meeting.
But I wonder if they really need two slots, especially when there is
also the Friday meeting.

> But of course I can be available for a presentation at the Global Identifier
> Task Force meeting as well.
> 
> In terms of the topic, I am not sure what exactly would be covered under:
> "Impact of global identifiers for dynamic reservation systems"
> 
> All the dynamic reservations systems currently use a global ID I believe. So
> perhaps a discussion would be "what is the impact to dynamic reservations
> systems if they were to all shift to the scheme proposed by our paper? Also
> what is the impact to E2EMON?"
> 
> Another topic would be, how do management/monitor systems (i.e. perfsonar)
> plan to use the global identifier as part its processes and functions?
> Thomas would probably be best to talk to that.
> 
> Also, I did not see a slot during the "Technical and Control Plane Working
> Group" or "Plenary" session where you would have a chance to talk about
> Global ID scheme. There is this slot:
> "15:50-16:30  Reports from Working Group and Task Force meetings", but based
> on time, will only be an opportunity to say that there is a paper to review.
> Just wondering how we get review comments, and eventual approval from GLIF
> as a whole.
> 
> Just some thoughts for further discussion at this point.

Good points. I think Global Identifiers will be used by the network
engineers to identify lightpaths in case of outages and announcements
of planned work. Idealy, this will be the case for both manual and
dynamic lightpaths. For manual provisioning, the global identifier
can be communicated during the provisioning phase.

For dynamic lightpaths, this is much more difficult, I think. You say
the dynamic provisioning systems use global identifiers. I think these
must be coupled to true unique global identifiers using our urn scheme
(or whatever we decide upon). That's the only way to make sure that
outages and planned work can be handled correctly. So I guess the
provisiong systems need to pass the urn global identifier to each
other and there has to be a way by which the network operators can
query the dynamic provisioning systems for the urn global identifier
of a particular lightpath.

I agree we need to right audience for these discussions. Having the
GNI clash is unfortunate. Do you have contacts with those people?
Would it be possible to squeeze the GNI meeting in the Monday slot
only?

	rvdp

Attachment: signature.asc
Description: Digital signature