Multi-WAN with IPv6: PPPoE DHCPv6 stuck offline after WAN2 reboot
-
A strange situation has been discovered with the combination of multi-WAN and IPv6 when IPv6 is enabled on both interfaces.
For example, there are two WAN interfaces: one is just WAN PPPoE, and the second is WAN2 DHCP + WAN2 IPv6 SLAAC for IPv6. On top of that, everything is running in CARP. In CARP, naturally, only WAN2 is present — PPPoE is not involved, except for a script that in this case does not react to events. Everything seems to work normally. I set the default interface for IPv6 to be the DHCPv6 one on PPPoE.But here’s the nuance: sometimes the upstream router that provides WAN2 reboots, and then some mess starts appearing in the logs. For some reason, the PPPoE DHCPv6 address gets “flapped,” and it ends up with WAN2 recovering normally, but PPPoE DHCPv6 stuck in offline or pending status don't remember exactly. If I enable the option disable monitoring actions (or whatever it’s exactly called) in the pfSense routes settings on pppoe dhcpv6, then PPPoE DHCPv6 continues functioning normally. Ipv4 pppoe looks unaffected, at least it continues to work normally in both situations.
Maybe I just misconfigured something, or is this kind of setup simply not supposed to be used?
WAN2 IPv6 is used exclusively by pfSense itself for updates and occasional debugging.
I can provide some logs later.
-
