Re: draft-ietf-v6ops-nap-00.txt <PROXIES>

[Pekka Savola <pekkas@netcore.fi>]

On Mon, 4 Apr 2005, Mark Smith wrote:
> > In any case, proxies become a necessity when ULAs are used.  The
> > situation differs because in v4 you can use NAT ("implicit proxy")
> > instead.
>
> I've understood that the intended, most common use of ULAs would be to
> deploy them in parallel with globals, rather than instead of globals. I
> wouldn't think a proxy would be a necessity in that scenario. Or have
> I misunderstood what you're saying ?

That's correct, but I suspect the situation may be different in some 
cases.  There will certainly be large deployments which only have ULA 
addresses, I suspect.

> In a ULA only scenario, a proxy would be necessary to access external
> content. It's use as a topology hiding mechanism wouldn't be of much
> value though, as the ULA addressed topology would be unreachable over
> the public Internet anyway.

The users who wish to hide the topology likely also want to hide 
non-routable topology.  Thus, using a proxy is benefit because by 
definition the local topology (whether non-routable or not) cannot be 
discerned.

As said, using proxies in ULA-only case (note that it could be that 
some nodes are ULA-only [but still need to have controlled access out] 
and some may be ULA+global) might fulfill some other perceived 
requirements rather than just topology hiding -- e.g., NAT-like 
security.

--
Pekka Savola                 "You each name yourselves king, yet the
Netcore Oy                    kingdom bleeds."
Systems. Networks. Security. -- George R.R. Martin: A Clash of Kings