Hi Bo,
The discussion slightly digressed. But, its fine and for this comment,
its
not really true that host translation will preserve end-to-end
transparency.
That is a myth. Any time Hui's PNAT host with its IPv6 transport
communicates with an IPv4 host on the internet, Hui will move the IPv6
packet translated from the IPv4 packet to the NAT64 gateway (to be
precise,
PNAT64) anyways, and there the IPv6 to IPv4 packet translation will be
done.
In fact, per Hui its only header translation on the host and the payload
translation is done on the CGN gateway. So, there goes the end-to-end
transparency and the simplified network argument.
[Bo] If in 4-6-4 scenario PNAT goes the E2E transparency argument. I believe
DS-lite will goes the argument as well. DS-Lite need the AFTR has the
capability of 44 translation, also against the concept of E2E transparency.
Agree. I dont believe DS-lite makes E2E transparency argument. So, we
both agree, host-based translation does not have E2E advantage in
this case.
> Disadvantage: it improve the UE cost explicitly, and others.
>
And exponentially, consider all the OS's and version variants. But, on a
lighter note, there is one more advantage point that you missed,
employment
for many folks for years to come for managing the PNAT stack on many
OS's :)
[Bo] If it is true, it is really a good solution not only for the company but
also for the global economy crisis.
:)
Regards
Sri
_______________________________________________
Softwires mailing list
[email protected]
https://www.ietf.org/mailman/listinfo/softwires
