Weird behavior with CARP failover

[Hidigoudi]

Hello,

I'm encountering an issue with 2 OPNsense instances running as VMs (on Proxmox) during failover from one router to the other. I've configured both OPNsense systems to replicate and allow a seamless failover from the master to the backup. However, every other time, the failover doesn't work properly and I lose connectivity on my WAN VIPs (public IP) for more than a minute (I just run a continuous ping). For the connection to come back, I have to leave CARP failover so that the original router becomes the master again.

Sometimes, the failover works perfectly and I only lose one or two pings. It is completely a random issue.

It seems like the states of the second nodes are not synchronized correctly.

The network cards are exactly the same on both OPNsense (same configuration). The first OPNsense is master and the second is SLAVE as expected so they can communicate each other.

The dark theme screenshots are referred to the MASTER node, the white theme is for SLAVE one.

Is someone able to help me ?

Thanks.

[Hidigoudi]

Others screenshots (I couldn't post more on my first message)