[CentOS] Configuring ipv6 reboot persistence, CentOS 5.6

Sat May 7 21:54:29 UTC 2011
Stephen Harris <lists at spuddy.org>

On Sat, May 07, 2011 at 04:28:45PM -0400, David Mehler wrote:
> Testing that with a ping6 works fine. I then want it to persist across
> reboots. So I added the following to /etc/sysconfig/network:
>        NETWORKING_IPV6=yes
>        IPV6_DEFAULTDEV=sit1

Looks good; I have the same.

> and I made /etc/sysconfig/network-scripts/ifcfg-sit1
>        DEVICE=sit1
>        BOOTPROTO=none
>        ONBOOT=yes
>        IPV6INIT=yes
>        IPV6TUNNELIPV4=IPV4 Address
>        IPV6ADDR=IPV6 Address

I have

> Reboot the box, check sit1 and it has an ip address. Running that
> ping6 command says the network can't be reached. I have to take the
> interface down and rerun those above commands manually. I'd appreciate
> any suggestions.

Is the IPv4 address in the config above the IP address of the HE endpoint
and _not_ your IP address?  (In my case I have for tserv3)

ifconfig sit1 should show something like
  sit1      Link encap:IPv6-in-IPv4
            inet6 addr: your_ip6/metmask Scope:Global
            inet6 addr: fe80::link_addr/64 Scope:Link
            UP POINTOPOINT RUNNING NOARP  MTU:1480  Metric:1
            RX packets:756 errors:0 dropped:0 overruns:0 frame:0
            TX packets:758 errors:0 dropped:0 overruns:0 carrier:0
            collisions:0 txqueuelen:0
            RX bytes:76421 (74.6 KiB)  TX bytes:88155 (86.0 KiB)

Also check the output of
  ip -6 route | grep -v 'dev lo'