dpinger defunct after adding new v6 gateway

[schmuessla]

Hi all,

I currently have a Multi WAN Setup with WAN1 v4/v6 + WAN2 v4 running.
I want to play around with v6 Failover. However if I add v6 configuration to WAN2
the WAN1 v6 stops working (not running) and the gateway is being marked as down, which is not the case because the watch IP is still pingable from opnsense's shell.

So far I couldn't find any log which gives hints why dpinger doesn't start anymore. Where can I find it?

[schmuessla]

Does somebody know where to start to dig into this, please?

[schmuessla]

I've dug a bit more into it. When I try to start dpinger manually with the command opnsense would have generated it immediately exits with

Code Select Expand


bind: Can't assign requested address
cannot bind send socket

also

Code Select Expand


ping -6 -s <WAN_DHCP6 address>


fails instantly with the same error. As soon as I remove the v6 configuration from WAN2 both work again.

This looks like a kernel bug to me? Or is it possible that this is caused by a invalid configuration? I tried both DHCPv6 and SLAAC on the WAN2 interface (a mobile router is connected to that)

[markus.tobatz]

See my post at https://forum.opnsense.org/index.php?topic=39995.msg196217#msg196217
Manually re-attaching the IP address via ifconfig command solves the issue for me. Sure, that doesn't explain the reason and can't be a permanent solution (i.e. I get a dynamic prefix from ISP, so I can't hardcode it on boot).