RE: ocean: do not boil

["Bound, Jim" <Jim.Bound@hp.com>]

yes.  this would be a cost for a network wanting to do this they would have to weigh against other approaches.  This is the kind of metric I think we need to record for each mechanism.

/jim

> -----Original Message-----
> From: Erik Nordmark [mailto:Erik.Nordmark@sun.com]
> Sent: Thursday, September 26, 2002 10:16 AM
> To: Keith Moore
> Cc: Margaret Wasserman; Bound, Jim; itojun@iijlab.net; 
> Stewart Tansley;
> v6ops@ops.ietf.org
> Subject: Re: ocean: do not boil 
> 
> 
> > it's a tradeoff - NAT-PT is probably slightly less 
> functional to apps (will 
> > cause a few more apps to break than NATs do) but in 
> exchange for using it
> > you get to get rid of all IPv4 traffic on your local network.
> 
> But you can get rid of the local IPv4 traffic (i.e. no need 
> to configure 
> local IPv4 routing) by tunneling IPv4 in IPv6 locally and 
> using the DSTM
> DHCP options for getting such tunnels configured.
> 
> This requires implementing more stuff on the nodes, and it adds
> tunneling overhead on the wire.
> 
>   Erik
> 
>