Print

[microtech]

Hi,

My current IPv4 setup has a linux router / firewall and various windows/linux clients behind that.

I've got the tunnel up and running on the linux firewall and can ping ipv6 stuff over the internet.

Can anyone point me in the right direction for instructions on how to configure windows and linux clients behind the firewall to route their own ipv6 public ip via the linux firewall?

I have curently requested the routed / 48 and have tried assigning ip's from that to the windows machine with the gateway set to the 'Client IPv6 address:', but it's doesn't seem to work.

I have iptables and an NAT'ing a private IPv4 space out via the router at the moment.

Any help for an ipv6 newbie would be apreciated.

Thanks.

[jimb]

You don't need a routed /48 unless you have more than one subnet behind your linux router.  The routed /64 should be sufficient.

You would use your routed /64, or a /64 subnet of your routed /48 on the LAN behind your router.  You should assign your IPv6 router (linux box) an IPv6 out of the /64 on the LAN interface (e.g. 2001:db8:1234::1/64).

The easiest way to get it going on the LAN is to use radvd, which is a daemon which performs Route Advertisement.  You can use this to advertise your /64 IPv6 prefix to your LAN, and all the machines on the LAN will autoconfigure.  You could also use DHCPv6 but not all OS support this natively and would need client software.

You could also statically assign IPv6 addresses as you're trying to do now.  Just assign them out of the routed /64.  The default gateway should be set to either the public IPv6 you put on the router's LAN interface, or the Link Local address of that same LAN interface (btw, if you use radvd, this will actually advertise the LL address).

You should also look at ip6tables on your router to make sure that is configured correctly.