Re: AD review: draft-ietf-kink-kink [section 1-4]

> [**] Discuss status message, rebooting peers and u2u.  This looks a
> lot like the IKE case where you lose all cryptographic context to me.

When the responder reboots, and an initiator sends a CREATE with
USE-SESSION-KEY to the responder, then the responder uses

	1) a identical TGT like storing a flush card.
	2) a different TGT from the previous, like using PK-INIT

to decode the message from the initiator.

#1. rebooting does not matter at the both sides.

#2. the responder can not decode the message, because the responder
does not have the previous session key at all.  my proposal is that
the responder sends back KRB_TGT_REP including its new TGT..(A).
then, the initiator can proceed to the next step like GETTGT operation.

The initiator MUST NOT believe that the responder rebooted by
receiving the KRB_TGT_REP message.