[ppml] Random v6 discussions (was Re: Policy Proposal: IPv4Transfer Policy Proposal)

[Ted Mittelstaedt]

>-----Original Message-----
>From: ppml-bounces at arin.net [mailto:ppml-bounces at arin.net]On Behalf Of
>Kevin Kargel
>Sent: Thursday, February 14, 2008 11:15 AM
>To: ppml at arin.net
>Subject: Re: [ppml] Random v6 discussions (was Re: Policy Proposal:
>IPv4Transfer Policy Proposal)
>
>
>
>> In our own experience, we've found that dual-stacking the 
>> edge is more difficult than doing the core or border because 
>> the edge equipment in the network you control is relatively 
>> less likely to be IPv6 enabled or upgradable for a variety of reasons.
>> 
>Ah, I am lucky enough to have Cisco network hardware throughout the
>enterprise..  perhaps that makes things easier.  I realize that not
>everyone has budget or desire to do that.
>
>It will at the moment put the onus on our customers to use bridged
>modems and either a software router or perhaps one of the two two
>available small hardware routers that support IPv6.  

Another possible option is distributing RFC1918 IPv4 to your end
users and dual-numbering your mailserver and a big web proxy server on
your network - just don't interconnect the
RFC1918 to the Internet with a NAT.  (If you have ever experienced
a NAT behind a NAT, it's dog-slow)  

Obviously this would only be good for remotes that just do e-mail
and web surfing.

Ted