On Nov 9, 2009, at 15:15, Brian E Carpenter wrote:
> On 2009-11-10 04:11, Chris Engel wrote:
>> 
>> 2) NAT serves to abstract internal hardware that provides services from the 
>> external advertisement of those services. Making it very easy and flexible 
>> to distribute and redistribute the provision of those services among your 
>> internal hardware.
> 
> I though we covered that, though not in those words. Certainly topology 
> hiding, one of the areas we discuss at some length,is a big part of that (and 
> is part of the gap analysis, so we tend to agree that is a weaker area).
> 
> Also of course ULAs are specifically designed to "abstract internal hardware 
> that provides services from the external advertisement".

Actually, I think Mr. Engels wants port-translating IPv6/NAT, expressly for 
making *surjective* mappings from unique local port/addresses to public 
port/addresses.  He doesn't seem to want this for topology hiding purposes, so 
much as for achieving high service availability.

I don't think I see anything in RFC 4864 that specifically addresses the usage 
of port-translating NAT as a tactic for achieving high service availability, 
and that might be the source of Mr. Engels' agitation.

Of course, there are perfectly reasonable ways to achieve high service 
availability without using NAT to make surjective address/port mappings, but 
RFC 4864 doesn't mention any of them.

Perhaps, an update *is* in order.


--
james woodyatt <[email protected]>
member of technical staff, communications engineering


_______________________________________________
nat66 mailing list
[email protected]
https://www.ietf.org/mailman/listinfo/nat66

Reply via email to