"Leasing" of space via non-connectivity providers

Fri Feb 11 03:28:40 UTC 2011

On 2/10/2011 9:11 PM, Jared Mauch wrote:
> I was explaining to my wife today how it felt like the nanog list went to 3x the typical mail volume recently with all the IPv6 stuff this month.  Why the pro-IPv6 crowd was happy, the anti-IPv6 crowd is groaning (including those that truly despise the whole thing, etc..)

I was having fun discussing with my wife how ARIN stuff ended up on 
NANOG, NANOG stuff ended up on PPML, and I've been listening and 
participating in debates concerning IPv6 and CGN (apparently BEHAVE WG 
adopted CGN over LSN) on 4 different mailing lists.

To be honest, though. I'm pro-IPv6, but I'm not happy. Anyone who is 
happy doesn't care about those innocent people who are ignorant of what 
is going on and why.

> I honestly think that the LSN situations won't be as bad as some of us think.  The big carriers have already been doing some flavor of this with their cellular/data networks.  Doing this on some of the consumer networks will likely not be "that much" pain.  Obviously the pain will vary per subscriber/home.

<snip lots of good stuff I agree with>
> IPv4 is "dead" in my opinion.  Not dead as in useless, but to the point where I don't think there is value in spending a lot of time worrying about the v4 side of the world when so much needs to be fixed in IPv6 land.
Service requirements in cellular networks are considerably different 
than wireline. Apparently, most cell customers don't hook a CPE router 
into their cell network and play their game consoles over it, along with 
many other situations. This actually means that most often, they are 
running a single stage NAT44 LSN (which still breaks stuff, but most of 
the things it would break aren't normally transiting the cellular networks).

<snip more good stuff I agree with>

I agree. However, because the largest networks and corporations decided 
(and some still do) to wait until the last moment to deal with IPv6, we 
will have to deal with IPv4 in much worse conditions. I know that there 
are large cellular networks which use DoD bogons behind huge LSN 
implementations. I know that some networks apparently aren't happy with 
using DoD bogons and would like to waste even more space. The best 
solution for such a case (and to solve all arguments on the matter) is 
to secure assurances on the bogons so that they can be safely used.

Jack