[Nsi-wg] comments on NS Framework 0.3
jrv at internet2.edu
Wed Jul 7 08:14:27 CDT 2010
Hi Guy -
A couple comments below. I am guessing you haven't had time to look at the edits I sent - I am interested in your comments and questions.
On Jul 7, 2010, at 8:42 AM, Guy Roberts wrote:
> In response to your questions:
> 1. In version 0.4 I have added the following definition of a Network to section 2.6: The Network object is a grouping of STPs which belong to a single NSA. A Network may have an associated transfer function matrix between STPs.
This is a transport layer definition - which seems basically right. What is missing is how a network service in an NSA / service plane relates to this in the transport plane.
> 2. Yes, the sections that have been removed will go in the Connection Service Architecture. If this will be a single document with the Connection Service Protocol is not decided yet.
> 3. In one of the discussions Kudoh-san suggested leaving out formalizing the INJ object, this is what I have done. This leads to some ambiguity as in places the STP is described as interconnecting networks, and in other places as the concatenation point between connections. We need to decide if this is ok.
Yes - we discussed this on a call and agreed that having both STP and INJ (or connector as Vangelis' suggests) was good. As I recall Tomohiro was agreeable to this and Jerry thought we needed two different terms (and we might find a catchier term for INJ).
> -----Original Message-----
> From: John Vollbrecht [mailto:jrv at internet2.edu]
> Sent: 06 July 2010 19:25
> To: NSI WG
> Subject: [Nsi-wg] comments on NS Framework 0.3
> Sorry not to get these sooner, but we have had a holiday and I have had trouble finding time to go through it. I attach a copy with tracking on, that includes edits and comments. In general I think it is getting better - Some overall comments -
> I think the format of the document is getting better and simpler. The wording in most places is good, with some edits in the doc. There are a few conceptual questionss that I have still --
> 1. We agreed to define network and network service. I don't see that, or at least not a clear definition. Is that to be added?
> 2. I assume some stuff will be incorporated in the Connection Service Architecture -- is that still right?
> 3. The concept of STP as described is not consistent between sections. I believe that we agreed that an STP is a point in a transport net which can be connected to another STP in the same net by an NSA. I believe we agreed to use the INJ concept for the connection between STPs in different transport networks. This seems to have reverted to the STP and INJ being the same thing, with no entity that describes the edge of a network at a physical (resource) level.
More information about the nsi-wg