comment below On Sep 7, 2009, at 5:45 AM, Guy Roberts wrote:
Hi Jeroen,
Thanks for pointing this out - I have misunderstood the definition/ use of transitional links. But now I really have a problem seeing the point of NML using transitional links. Re-reading I wonder if this is perhaps a concept only required for George's path computation on a 'flattened multi-layer topology' idea? This is what he says about this:
"2.4 Transitional Links
The discussion above shows that use of the G.805 link as the topological entity between nodes in a routing graph is too restrictive. We need another term to describe the node – node relationship in the network graph. In graph theory the terms node and edge or vertex and arc are used, and there is value in adopting vertex and arc (or edge) to avoid confusion. Nodes frequently designate equipment.
Inspection of the figure suggests that the links that are important when considering pruning are those links whose CI changes along the link (L1, L2 and L3). L1 and L3 convert CI from the client to the server (and back again) while L2 converts server CI to client at each end. If we include those links, we have a connected graph at the client CI. If we remove those links then we have the server topology left over. It seems that we are perhaps better off retaining the term link, because it has been use so long, and introducing the term transitional link for those links which change CI along the link. This should give us some strong hints when looking for appropriate link semantics."
If path computation is the only proposed use for transitional links then I am not sure that we need to use the concept in NML, or are we proposing another use that I have missed?
I think that transitional link is a way of flattening topology where it is possible to do adaptations between layers. It requires that a device be represented at both layers and have a link between layers with a specific adaptation. When doing pathfinding using such links one must be sure adaptations and deadaptations match such that the client info is passed through. I think this concept is highly useful for describing a "routing area" over which pathfinding can be done. We should consider this relative to other methods of pathfinding which seem more difficult to use in practice. John
Guy
-----Original Message----- From: Jeroen van der Ham [mailto:vdham@uva.nl] Sent: 04 September 2009 19:13 To: Guy Roberts Cc: John Vollbrecht; NSI WG; Network Markup Language Working Group Subject: Re: [Nml-wg] [Nsi-wg] Conversation about ITU concepts with Ciena folks
Guy Roberts wrote:
The other document of note is the transitional link document. I think we need to be careful about adopting this concept since as far as I can see it has been created by ITU-T for a very specific purpose. They use it for transit between sub-layers as opposed to adaptation between layers. In their example a transitional link is used for all-optical conversion of wavelengths, where wavelengths are not real layers as there is no termination and adaptation function when converting between wavelengths.
That's not completely true. As far as I see it there are two kinds of Transitional Links. The example you are referring to are sub-layer transitional links, where the connection on a single layer is not terminated, and readapted, but only the "Characteristic Information" is transformed, in this case the wavelength.
There is also the Layer Transitional Link. This link does go from a client layer to a server layer, without specifying all the details. It seems to be an abstraction of a normal link with an adaptation.
I am not sure whether the distinction between sub-layer and layers is actually relevant to us. So far I'm inclined to think that it is not relevant.
Jeroen.