Re: [RAM] Re: the separation of ID/RLOC

[Tony Li <tli at cisco.com>]

On Jul 11, 2007, at 3:34 AM, Christian Vogt wrote:

> > If E prefers P1 as ingress and F prefers P2 as egress and there's no
> > P1-P2 link, then you end up violating E's preferences.

> why would E prefer an ingress provider (P1 in this case) via which it
> cannot reach correspondent edge networks?
>
> Or vice versa, why would F prefer an egress provider (P2 in this case)
> via which it cannot be reached by correspondent edge networks?


I was assuming reachability, otherwise the entire question makes no  
sense.


> You are raising an interesting point.  I wonder how often this  
> happens,
> though.  Do you have some more insight on this?  And one should also
> take a look at the impact of such asymmetry on applications...


Sorry, not much.  I'm sure that in the whole of the Internet today,  
there must be at least one case like this currently in existence.   
It's probably masked by the multihomed networks using a PI address  
instead of PA.

Tony

_______________________________________________
RAM mailing list
RAM at iab.org
https://www1.ietf.org/mailman/listinfo/ram