RE: [DNA] Issue about multiply connected to the same link and DNAbeing per-interface or not (was Re: WG LC comments ondraft-ietf-dna-goals-01.txt)

[Alper Yegin <alper.yegin@samsung.com>]

> > I was mostly objecting the notion that we should keep DNA on
> > separate interfaces _completely_ independent.  

That was not my intention when I said "DNA scheme is run per-interface".
What I meant was the result that is produced by DNA scheme (e.g.,
whether IP reconfig is necessary) must have a target interface. It
cannot be per host, which may have multiple interfaces. Getting to that
result may use input from other network interfaces, and even non-network
interfaces (e.g., GPS), that's fine. 

The current text refers to "host" changing links. Actually, it is a
network interface on the host that is changing links. A dual-interfaced
host may disconnect eth0 from link_x connect eth1 to link_x. Host is
still connected to the same link, hence no "link change per-host", yet
the IP configurations on the eth0 and eth1 has to change. 

Maybe we should say "DNA scheme is run per-interface. While the
detection is performed for an interface, the DNA scheme may take hints
from other interfaces when useful. "

Alper