2
 A transit network is represented by a graph vertex having both incoming and outgoing edges

I can't say that this representation makes sense for me. Can anybody explain why the OSPF algorithm needs this?

Here is an exaple enter image description here

On top of the above some edges in that graph seem to have different weights for the same link

3

It has two edges to represent each direction, and each direction of a generalised SPF could have a different path cost.

  • Thank you! In the end I got it too, what confused me was the fact that I was thinking about the cost of the link as being the speed of the link but ...the cost of the link might not have anything to do with that and it could be administratively set, and that could be different in each direction – MiniMe Jul 5 '16 at 13:55
  • well at a second thought I still have questions: if you have a broadcast segment then you have still have two edges connecting the node to the segment. I can understand two edges on a P2P link but why do you need two edges for representing a connection to the broadcast network – MiniMe Jul 5 '16 at 14:00
  • I don't really know because although I meant to code the SPF algorithm myself to fully understand it, I never actually got around to it. I guess that because of the possibility that a particular path may be uni-directional you write the path cost algorithm to follow outbound edges, so even if you don't have a cost associated with "leaving" a broadcast network, you must still have an edge to follow. – marctxk Jul 5 '16 at 14:11
  • To preserve the two vertices model of P2P link without impacting the calculation. Here is the explanation: "The concept behind the DR is that the broadcast link itself is considered a "pseudonode," or a virtual router. When the SPF tree is calculated, the link appears as a node and the routers attached to the link are attached to that node. The cost from an attached router to the pseudonode is the outgoing cost of that router's interface to the broadcast link, but the cost from the pseudonode to any attached router is 0. This way, the overall path cost is not affected by the pseudonode." – MiniMe Jul 5 '16 at 14:27
4

Here is a clarification to marc's answer

"If you have a broadcast segment then you have still have two edges connecting the node to the segment. I can understand two edges on a P2P link but why do you need two edges for representing a connection to the broadcast network "

In order preserve the two vertices model of P2P link without impacting the calculation. Here is the explanation: "The concept behind the DR is that the broadcast link itself is considered a "pseudonode," or a virtual router. When the SPF tree is calculated, the link appears as a node and the routers attached to the link are attached to that node. The cost from an attached router to the pseudonode is the outgoing cost of that router's interface to the broadcast link, but the cost from the pseudonode to any attached router is 0. This way, the overall path cost is not affected by the pseudonode." – MiniMe just now edit

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.