The IP Over NBMA (ION) Archive

[Masataka Ohta <mohta@necom830.hpcl.titech.ac.jp>]

> >> Then, as the Path state flips, the cell loop is formed and released
> >> and looping data will be discarded.
> 
> It is *high possibility* that the cell loop is discarded.

What are you worrying about?

The point is NOT that CSRs are more loop-resistant than normal packet
routers. CSRs are not. CSRs are just as loop-resistant as normal
packet routers.

> Because RESV message is gone in the loop, when R1 send RESV message
> toward src .

As long as routing loops exist, there is some possibility that
something loops continuously or continually, regardless of whether
they  are CSRs or normal packet routers.

So, the point is that, if routing restabilizes, the loops disappear,
which is obvious with normal packet routers or CSRs.

The problem is that NHRP may create a persistent loop even after
the restabilization, which has nothing to do with CSRs.

But, even it is not a real problem, because NHRP must disappear with
other reasons.

Can't you understand it?

							Masataka Ohta