IPv6 / 6to4 with a WRVS4400nV2

[DominikAu]

Hi,

i've been messing around with the configuration since hours, but failed
to get a stable IPv6 connection into the internet. While doing that,
i've also discovered several shortcomings.

First things first, the configuration of the router:
WAN -> Static IP
IP-Mode -> dual stack
LAN -> DHCP for IPv4 and IPv6 activated
DHCPv6 setup:
    Router Advertisement: activated
    IPv6 Prefix (generated by the router, but i've checked it, it
matches the 6to4 notation of my public ip) , IPv6 Postfix is :1, the
Prefix Length 64
    Lease time 360
    DHCP6 address range start: 2002:"public-ip"::10
    DHCP6 address range end: 2002:"public-ip"::99
    Primary DNS: 2001:4f8:0:2::14

Settings on a client (windows 7):
    2x IPv6-adresses with the correct prefix (out of the router
advertisement)
    default gateway: fe80:: ... (link local address of the WRVS4400)

if i try to ping www.six.heise.de (2a02:2e0:3fe:100::6),
ipv6.google.com, ... i get four ICMPv6 replies from the router, which
are of type 1 code 3 (Destination Unreachable, Address unreachable).

But there is a very small amount of ip's which are ping-able, which are,
2002:d596:2a92:1:71:53:: and 2002:5968:c28e::53. So the basic connection
to the anycast-ip 192.88.99.1 should work. Further, the "unreachable"
replies come immediately, so it seems, that the router don't even tries
to reach the addresses.

As far as i've understood 6to4 should be able to connect IPv6 islands
via IPv4 connections, so i should also be able to access the "real IPv6
internet".

Sometimes i get some Type 3 Code 0 (time exceeded, hop limit exceeded in
transit) responses from the router before failing with the unreachable
replies.

Further i've discovered that the DHCPv6 server is pretty unstable, it
just stopps to respond  to solicitations a few minutes after a reboot.
In addition it won't even start when using a /48 subnet. In that case
there is a log entry which states that the prefix length at br0 should
be 64, which isn't true for 6to4 addresses.

I've already reset the router by holding the reset button three times
for more than 30 seconds.

Kind regards,
Dominik Augustin

[DominikAu]

it seems that this doesn't happen to other ones.

so i suspect that there's a hardware bug /defect memory in my unit.

should i just send it back to my seller?

[DominikAu]

the problem is hidden in the routing table of the router.

since the (ipv6) settings aren't accessable through the WebUI, i had to downgrade to a older firmware, get a utelnetd from a rvs opensource firmware package and copy it onto the router. that done, i was able to activate the telnet server through the [routerip]/Hidden_telnet.htm page and access it the usual way (putty / windows / whatever).

then i explored the ip -6 route table and found

2000::/3 dev sit1  metric 1

so there is no via and no way

aren't you (cisco) the network guys who should know such basics? (since i'm no linux or network guru, this one should have been obvious to you)

change it with

ip -6 route del 2000::/3

ip -6 route add 2000::/3 via ::192.88.99.1 dev sit1

and tadaa i'm able to ping (and reach) the whole ipv6 internet.

btw: if the router allows downloading the utelnetd, the route change should be possible throught the gui. but since cisco loves to close such things instead of removing bugs, ...

btw:

ip -6 route also produced:

unreachable default dev lo  metric -1  error -101