I have setup OpenVPN and now need to setup the routes manually, but I cannot manage to ping the other end. The tun0 interface is up and has the IPv6 address fd2d:65e3:ed0d:8::1010/64 (global scope) set by OpenVPN. Now I try to reach the other end.
$ping6 -I tun0 ff02::2
PING ff02::2(ff02::2) from fd2d:65e3:ed0d:8::1010 tun0: 56 data bytes
64 bytes from fd2d:65e3:ed0d:8::1010: icmp_seq=1 ttl=64 time=0.049 ms
64 bytes from fd2d:65e3:ed0d:8::: icmp_seq=1 ttl=64 time=63.2 ms (DUP!)
The other sides IP address is fd2d:65e3:ed0d:8::, alright, let's first add a route:
$ip route add fd2d:65e3:ed0d:8::/64 dev tun0
.. and now let's ping the other side:
$ping6 fd2d:65e3:ed0d:8::
PING fd2d:65e3:ed0d:8::(fd2d:65e3:ed0d:8::) 56 data bytes
64 bytes from fd2d:65e3:ed0d:8::1010: icmp_seq=1 ttl=64 time=0.031 ms
Why is the local interface responding? How can I ping the other end of the tunnel?
EDIT:
ip route add 2000::/3 via fd2d:65e3:ed0d:8:: dev tun0
... does get me "RTNETLINK answers: No route to host", but this is allowed:
ip route add 2000::/3 via fd2d:65e3:ed0d:8::1 dev tun0
Jeremy is about right. It is an anycast address. Since it is not really meant to ping/connect to the other side, the right thing to do is to not use the address at all:
Now we could e.g. ping google: