IPv6 addresses got via DHCPv6 Server add up and don't vanish after new lease is issued.

4920441 0

Hi,

I have got a problem with the change of the IPv6 Addresses assignet via DHCPv6 client on the current 2.4.5-RELEASE version.

The German Telekom provides dynamic IPv6 addresses and /56 prefixes.
I got two "uplinks", both to 'Deutsche Telekom' - the first one is directly connected by pppoe to the pfsense Box.
The second one is connected via a fritzbox.

The direct via pppoe connected uplink has no problem.

The second IPv6 Connection, where get a public (though dynamic...) IPv6 Address works fine, as long as it does not change. (I am lucky changing occurs only after 6 month or so, or if you manually reconnect)

But nevertheless, if the connection is reconnected, the DHCPv6 Server of the Fritzbox propagates the new IPv6 Address accordingly within seconds, but in pfsense the old Address is neither removed, nor is it timed to be removed in a couple of seconds

Befor new IPv6 address is pushed to pfsense via dhcpv6

lagg0.30: flags=8943<UP,BROADCAST,RUNNING,PROMISC,SIMPLEX,MULTICAST> metric 0 mtu 1500
	options=600703<RXCSUM,TXCSUM,TSO4,TSO6,LRO,RXCSUM_IPV6,TXCSUM_IPV6>
	ether 00:11:22:22:33:33
	inet6 fe80::222:9222:2222:2222%lagg0.30 prefixlen 64 scopeid 0xb
	inet6 2003:d0:aaaa:bbbb:bbb:bbbb:cccc:cccc prefixlen 64 autoconf
	inet 192.168.49.250 netmask 0xffffff00 broadcast 192.168.49.255
	nd6 options=23<PERFORMNUD,ACCEPT_RTADV,AUTO_LINKLOCAL>
	media: Ethernet autoselect
	status: active
	vlan: 30 vlanpcp: 0 parent interface: lagg0
	groups: vlan

After new IPv6 address is pushed to pfsense via dhcpv6:

lagg0.30: flags=8943<UP,BROADCAST,RUNNING,PROMISC,SIMPLEX,MULTICAST> metric 0 mtu 1500
	options=600703<RXCSUM,TXCSUM,TSO4,TSO6,LRO,RXCSUM_IPV6,TXCSUM_IPV6>
	ether 00:11:22:22:33:33
	inet6 fe80::222:9222:2222:2222%lagg0.30 prefixlen 64 scopeid 0xb
	inet6 2003:d0:aaaa:bbbb:bbb:bbbb:cccc:cccc prefixlen 64 autoconf
	inet6 2003:d0:cccc:dddd:eee:ffff:cccc:cccc prefixlen 64 autoconf
	inet 192.168.49.250 netmask 0xffffff00 broadcast 192.168.49.255
	nd6 options=23<PERFORMNUD,ACCEPT_RTADV,AUTO_LINKLOCAL>
	media: Ethernet autoselect
	status: active
	vlan: 30 vlanpcp: 0 parent interface: lagg0
	groups: vlan

I waited (accidentially) a whole day but the old address stood put in place, so far.

Problems resulting from that:

dyndns is not updateing correctly, because the other Address seems still unchanged
the old address is certainly unreachable...
and maybe some more I did not encountered yet.

The only workaround so far is to use the "release DHCP" option in the interface status section.
After that i simply got the correct singe IPv6 address and dyndns uses the correct one again...

(BTW: is it possible to initiate the release the exact same way as it is done in the webinterface in the console as well?)

Is this a bug or a feature or what can I do about it?

Thanks a lot!

Cheers

4920441