Re: issue 34 proposal

Francis Dupont <[email protected]> Tue, 20 Sep 2005 13:11:45 +0200

Newsgroups	gmane.ietf.mobike
Message-ID	<[email protected]>

 In your previous mail you wrote:

   Mohan Parthasarathy wrote:

   >I am assuming NAT change at ESP layer follows IKEv2 NAT-T i.e no
   >return routabilitycheck. Otherwise, the speedup obtained by the
   >traditional IKEv2 NAT-T is lost and there is no point in having
   >this mechanism at all.

   I think most of the speedup is in the detection part.

=> I believe you mean the detection is more efficient because there are
more ESP traffic, because the return routability check cost is one
full RTT (without RR check the delay is the mean time between two ESP
packets + 1/2 RTT).

Regards

[email protected]