Re: Proposed Resolution of Issues [1-37]

[Tim Chown <tjc@ecs.soton.ac.uk>]

On Thu, Sep 01, 2005 at 12:45:29PM +0200, Gunter Van de Velde (gvandeve) wrote:
> The issue list available on http://www.vandevelde.cc/ietf/NAP-Issue-Log.htm 
> has
> been updated and addresses the remainder of the issues from the WGLC.
> 
> The Consensus answers in marked in blue are the onces that were proposed
> last week and have been considered as stable consensus. The proposed 
> resolution
> entries marked in red are the onces added today.

Looks good.

For issue 5, RFC3041 might hamper that.

On Issue 9, I suspect Kurtis has good figures on this, I recall he did 
back before multi6 reformed and we had multihoming discussions in Michel's
list.  

Issue 10, because admins pick 10.0.0.0/24 or 192.168.0.0/24.  I'll wager
they do the same with ULAs still, but hey :)

Issue 12 ULAs dont aggregate internally either, as each /48 is in effect
a random prefix.  May bug big organisations, who thus may not use 'proper'
ULAs?

Issue 14 - as Stig pointed out, if two remote sites use ULAs they may
prfer to use ULAs between them.   But that could be configured as 3484
policy too.   The key thing is that ULAs now have global scope, so are
by default treated as globals.

Issue 20 - I'm very wary that we should not embed 'best practice' in
firewall configuration in this draft.  I'd say 'less is more' here.

Issue 23 - unique but may change, eg. 3041.  I see little difference from
the ISP view in a customer with a stable single v4 address + NAT and a
customer with a stable /56 prefix and 3041 in use.  The v4 address masks
systems just as the /56 in effect does.

Issue 24 - so what text will you use? :)

Issue 27... hmmmm.  Renumbering complexity has many variables.  A simple
home network with one LAN and no servers/services is as easy as it gets.
I'd assume that most current NAT-using sites don't have many externally
advertised services, but will embed IP addresses in their internal systems
and networks.

Issues 32&33:  ULAs *may* help renumbering, not *will*.  There is baggage 
with ULAs and as such their use is a tradeoff not a given, I feel.

Issue 36: in practice sites will run IPv6 with IPv4+NAT, so they need
to secure two protocols 'on the wire'.

Issue 37: noone is mandating a migration from IPv4+NAT to IPv6.  Indeed
many sites may equally benefit from a migration to use global IPv4 addresses.
One might even consider in the draft how much is IPv6 specific, and how
much is 'use global addresses' specific.

I'll also resubmit the posrt-scanning text soon.

-- 
Tim/::1