I have ordered a host with a major hosting provider and found IPv6 configured in a slightly unusual way.
The provider have allocated the prefix 2001:db8:469e:1ea1::/64
to my host. On the provider router (which is outside of my control) the prefix is configured as a routed prefix, so all IPv6 addresses in the prefix are routed to the MAC address of my host without any neighbor discovery taking place for each address.
On the host itself they have configured the same prefix as a link prefix:
# ip -6 address show eth0
2: eth0: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 state UP qlen 1000
inet6 2001:db8:469e:1ea1::1/64 scope global deprecated
valid_lft forever preferred_lft 0sec
inet6 fe80::9400:ff:fe85:6c57/64 scope link
valid_lft forever preferred_lft forever
Though this is a bit unusual I don't immediately see how it could cause any problems. The purpose of receiving a routed /64
is that you can configure that as a link prefix on another interface. And as such I don't think this is necessarily an incorrect configuration.
What I would have expected as a more standard configuration was that the link prefix and routed prefix were different /64
where I can use one address in the link prefix and all addresses in the routed prefix. I would have expected something like 2001:db8:469e:1e00:9400:ff:fe85:6c57/64
in the output of the above command. The routed prefix I would have expected either configured as 2001:db8:469e:1ea1::1/64
on a dummy interface or left for myself to configure.
Is this configuration from the hosting provider's side sane, or can it break in subtle ways that I haven't realized?
0 Answers