[serverbros]

Cheers neddy Sounds like OVH all over right enough .....maybe it'll be correct when the IPV6 pool is depleted in the year 2913

[NeddySeagoon]

Heres a wee snippet from my setup

config_eth0="5.135.178.112/24 brd 5.135.178.255
             2001:41d0:8:b970::2/56"
# we only have a /64 but need to use /56 here so the router can be reached
# otherwise the default route is out of our network

routes_eth0="default via 5.135.178.254
             default via 2001:41d0:8:b9ff:ff:ff:ff:fd"
#The router (default gateway) for each IPv6 is always on IP:v:6:FF:FF:FF:FF:FF

The key lines there are the

              default via 2001:41d0:8:b9ff:ff:ff:ff:fd"
#The router (default gateway) for each IPv6 is always on IP:v:6:FF:FF:FF:FF:FF

The coment is how OVH document the IPv6 gateway. What works, ending in fd, not ff, you get from a router discover message.
OVH are not interested in fixing this, or the documentation.

[serverbros]

Hmm, now with the default gateway set the ping6 does not work/return anything at all via the container

Back to the drawing board!

[serverbros]

Thanks Myatu I'll take a look tonight!

[Myatu]

So when setting up the pool I've put : 2607:5300:60:3FF:FF:FF:FF:FF for the gateway and generated 50 Ip's, one of them being 2607:5300:60:3a2f:0000:0000:0134:020b

The gateway would be:

2607:5300:60:3aff:ff:ff:ff:ff

Also:

• check that for testing purposes the ip6tables is empty on the host (ip6tables -F)
• Ensure that there's a route from your host and OpenVZ VM to the gateway. The gateway sits on a /56 netmask, but your IP range is on a /64 netmask, therefore the gateway fall outside the normal range - in other words, you need to specifically tell your system how to reach the gateway. Relevant portion in this help document is "route -A inet6 add default gw ...".

[serverbros]

Hi Folks,

Going to post this here in the offchance someone is so kind to help, Originally sent to OVH for help however they understandably can't help due to unmanaged, althought kind sir Mr Neil sent a few links which were helpful on this forum unfortunately for KVM though

My setup :

MSP in both Roubaix France & BHS Canada

Each have a /64 ipv6 routed and both respond as expected by ping6 from the main nodes, when assigning ipv6 to a openvz container we unortunately get the error shown below.

BTW Virtualizor is our control panel :

Hi Team,

I'm facing an issue on our node in BHS with IPV6 in that the IPV6 works on the host node with ping6, replies fine however on openvz vm's (just routed so should work fine) I get the following reply from ping6 :

PING ipv6.google.com(iad23s06-in-x14.1e100.net) 56 data bytes
From 2607:5300:60:3a2f::1 icmp_seq=1 Destination unreachable: Administratively prohibited

I've followed the guide which states "The IPv6 server: 2001:41D0:1:46e::/64 to 2001:41D0:1:4 + 5x FF.
IPv6 Gateway: 2001:41D0:1:4FF:FF:FF:FF:FF"

So when setting up the pool I've put : 2607:5300:60:3FF:FF:FF:FF:FF for the gateway and generated 50 Ip's, one of them being 2607:5300:60:3a2f:0000:0000:0134:020b

Am I doing everything correct, or have I failed to grasp IPV6 like the scottish **** I am ?

Thanks,
Jordan