[Hipsec] Using STUN for end-to-end connectivity checks

[Gonzalo Camarillo <Gonzalo.Camarillo at ericsson.com>]

Folks,

as I said in our WG session, I have pushed the NAT Traversal for HIP 
design team to make a decision on the protocol to implement end-to-end 
connectivity checks. The design team was spending too much time thinking 
of this issue and, in my opinion (as HIP WG chair), this was keeping the 
team from making progress.

There were two proposals to implement the end-to-end connectivity 
checks: using STUN or using HIP. The design team agreed to make a 
decision before the end of the IETF. Today, we met for two hours to make 
this decision.

The idea was to make a design assumption (i.e., to use STUN or to use 
HIP) so that the team can work on other open issues and implementers can 
start experimenting. If by the time the specification is more or less 
done and we have more implementation experience, we find overwhelming 
evidence that this was the wrong decision, the design team agreed that 
we would able to reevaluate the decision.

The design team could not reach unanimous consensus on which protocol to 
use because there were technical arguments in favor of both of them. 
Therefore, they had to vote. The STUN approach won the voting (4 votes 
for STUN  2 votes for HIP).

Consequently, from now on, the team will be working assuming STUN as the 
protocol to implement end-to-end connectivity checks.

Cheers,

Gonzalo
HIP WG co-chair

_______________________________________________
Hipsec mailing list
Hipsec at lists.ietf.org
https://www1.ietf.org/mailman/listinfo/hipsec