[Trust-router] Captured items from the side meeting

[Sam Hartman <[e-mail]>]

Hi.

I want to thank those who attended the side meeting and/or rtgarea
presentation at IETF 86.
I notice that we don't seem to have sent pointers to the bar bof slides
here.
I'll see if we can arrange for that soon.

I did want to capture a few items I heard at the side meeting and ask
for any items I missed.

1) Brian recommended that we keep the mapping layer (the routing
protocol) separate from the TID protocol so that another mapping
mechanism could potentially be substituted.
He drew an analogy between LISP and trust router.
In LISP, mapping between map servers and map resolvers is carried over a
BGP overlay in the default instantiation, but there are other proposals
for how mapping could be handled.

I certainly never noticed that similarity, and if Margaret did, she
never mentioned it to me. 
Thanks Brian for pointing that out; it's something to ponder.

2) Brian and Sandy both brought up the Byzantine failure question. I
hope I have provided an answer to that.


In particular, I'm wondering if any of Klaas or Hannes's questions
require updates or whether they made any suggestions for changes.

At the IETf meeting we also received a number of other use cases to
consider:

1) In the rtgarea meeting we were asked whether something like trust
router could key some of the KARP use cases; I think group keying.

2) We were advised to take a look at what's going on in CDNI and see if
their federated trust problem would be useful.

We're also aware from previous discussions that various AAA roaming
problems are similar to trust-router's problem statement outside the
specific ABFAB context.

--Sam