Re: Is a translator a router? [draft-ietf-behave-v6v4-xlate-05]

Brian Carpenter wrote:
> >> That is what my proposed text does (for the v4 to v6 direction):
> >>
> >>    If the translator is combined with an IPv6 router, the router
> MUST
> >>    perform normal Hop Count processing [RFC2460]. Otherwise the
> >>    translator MUST emulate this processing.
> >
> > The above text is confusing to me, since "combined" is ambiguous.
> > If a v4/v6 translator is running on the same box as a regular
> > IPv6 router, but they're not chained, just running ships in the
> night,
> > they're still "combined" in the "same box" sense.
> 
> Sure, "combined in series" is the case that concerns me. Specifically,
> I'm concerned if a perfectly OK forwarding engine (which decrements
> hop count) is combined in series with a perfectly OK translator (which
> decrements hop count).

I guess it doesn't particularly concern me.  It seems to be an internal 
implementation detail.  If someone really chooses to implement a 
single physical box that appears as two separate virtual hops,
one behind the other, it shouldn’t break anything.

-Dave
_______________________________________________
Behave mailing list
Behave <at> ietf.org
https://www.ietf.org/mailman/listinfo/behave