Can anyone explain this? Just after reboot.
Feb/02/2022 02:32:14 vrrp2 now BACKUP, got higher priority 100 from 10.200.1.254
Feb/02/2022 02:32:19 vrrp2 now MASTER, master down timer
Surely once VRRP has got a message from a higher priority device it should cancel the timer?
Then 12 mins later
Feb/02/2022 02:44:31 vrrp1 now MASTER, master down timer
Why does VRRP1 timer take longer than VRRP2?
Is VRRP just completely broken? Running 6.49.2
Post your config on both devices.
What does the 1 node say in the logs during these issues?
Nothing in the log of the other node besides initial msg about becoming master.
Node not reacting messages from the other:
/interface bridge
add name=bridge1 protocol-mode=none
/interface vrrp
add interface=bridge1 name=vrrp1 priority=80
/interface vlan
add interface=bridge1 name=data vlan-id=100
/interface vrrp
add interface=data name=vrrp2 priority=80 vrid=2
/interface bridge port
add bridge=bridge1 hw=no interface=ether1
add bridge=bridge1 interface=wlan-gateway
/ip address
add address=10.2.1.5/24 interface=ether1 network=10.2.1.0
add address=10.2.1.1 interface=vrrp1 network=10.2.1.1
add address=10.200.1.1 interface=vrrp2 network=10.200.1.1
add address=10.200.1.253/24 interface=data network=10.200.1.0
/ip firewall filter
add action=fasttrack-connection chain=forward connection-state=established,related protocol=!vrrp
add action=accept chain=forward connection-state=established,related
Node that sends the message that is ignored
/interface vrrp
add interface=wlan1 name=vrrp1
/interface vlan
add interface=wlan1 name=data vlan-id=100
/interface vrrp
add interface=data name=vrrp2 vrid=2
/ip address
add address=10.200.1.1 interface=vrrp2 network=10.200.1.0
add address=10.2.1.254/24 interface=wlan1 network=10.2.1.0
add address=10.200.1.254/24 interface=data network=10.200.1.0
add address=10.2.1.1 interface=vrrp1 network=10.2.1.1
Omitted stuff about wrireless etc
You may get it wrong, or I didn’t understand what you wrote.
In fact, the master down timer is restarted, not cancelled, by each reception of the VRRP message from a higher priority device. So the interface switches from BACKUP to MASTER whenever it stops receiving VRRP messages from a higher priority device for long enough.
I’d assume they are unrelated - it’s just that the VRRP messages start getting lost in VLAN 100 much sooner than in the (untagged on the wire/in the air) VLAN 1.
Did the backup router report vrrp1 now BACKUP, got higher priority 100 from 10.2.1.254 at least once?
Also, 10.2.1.5/24 has to be associated to bridge1, not to ether1, but that doesn’t explain the current behavior - unless ether1 is not a member port of bridge1.
OK you are correct the timer is restarted. But it isn’t! Because a few seconds later it times out and the node becomes master
| In fact, the master down timer is restarted, not cancelled, by each reception of the VRRP message from a higher priority device. So the interface switches from BACKUP to MASTER whenever it stops | receiving VRRP messages from a higher priority device for long enough.
How do you know what happens? To be sure, you have to sniff:
/tool sniffer quick interface=bridge1 mac-address=00:00:5E:00:01:01 (for vrrp 1) or mac-address=00:00:5E:00:01:02 (for vrrp2).
And watch the direction of the packets - if it shows → , the device where you sniff sends them; if it shows ← , the external device sends them.