HA A-P Preemption Issue: LACP interface stuck in 'mondev down' after reboot until manual flap
Hello everyone,
I am facing an issue with an HA Active-Passive setup using two FortiGate 120G devices (Firmware:7.6.6).
The Setup:
HA Mode: Active-Passive
Priority: FG1 (128) - intended Primary, FG2 (120) - intended Secondary.
HA Override: enable
Monitored Interfaces: ATC-LACP (802.3ad Aggregate interface connected to a core switch).
The Problem: When I reboot the Primary unit (FG1), failover happens correctly, and FG2 takes over. However, when FG1 finishes booting up, it does NOT preempt back to the Primary role. Running get system ha status shows that FG1 is stuck as Secondary with the following warning: WARNING: FG120GTKXXXXXXXX has mondev down;
The LACP interface on FG1 stays down. The only way to fix this and trigger preemption is to manually log into FG1 (which is currently the secondary) and flap the interface:
Plaintext
config system interface
edit "ATC-LACP"
set status down
set status up
next
end
Immediately after this manual flap, the LACP comes up, the mondev down warning disappears, and FG1 successfully preempts to become the Primary again.
What I have already tried: On the FortiGate side, I have configured the following on the LACP interface:
set lacp-speed fast
set link-up-delay 10000 (to give it 10 seconds after reboot)
set lacp-ha-secondary enable (to keep the interface active while secondary)
Despite these settings, the manual flap is still required after a reboot. The switch is currently using standard (slow) LACP timers, and I am waiting for the switch admin to change it to fast.
My Questions:
Is the switch's LACP rate (slow vs. fast mismatch) the sole reason the FortiGate's LACP interface fails to initialize automatically after a reboot?
Are there any other specific FortiOS CLI commands or best practices to ensure the LACP comes up automatically and triggers preemption without manual intervention?
Any advice or insights would be greatly appreciated!
Thanks in advance.