Technical Tip: Reboot behavior when enabling or disabling Enhanced load balancing for LAG interfaces for NP6 platforms

This article focuses on the reboot requirement when enabling or disabling Enhanced load balancing for LAG interfaces on NP6-based FortiGate models. While the feature improves traffic distribution across LAG member links, any change to this setting requires a system reboot.

When the command is applied, the system confirms that a reboot is required:

FG1 (global) # config system npu
FG1 (npu) # set lag-out-port-select enable
FG1 (npu) # end
All LAG algorithm will be changed to hiding npu.
The configuration will take effect after system reboot.
Do you want to continue? (y/n)y

Secondary will reboot first, followed by primary reboot.

• For standalone units, the device reboots immediately after confirmation.
• For HA clusters, the reboot is coordinated:
  • The secondary unit reboots first.
  • The primary unit waits until the secondary has come back online and taken over the primary role, or for a maximum of 3 minutes after it boots up, whichever occurs first.
  • The previous primary then reboots, completing the configuration change across the cluster.

Recommendations.

• Plan a maintenance window as a reboot is mandatory.
• Verify model support: Available only on NP6-based FortiGates with ISF support.
• Check HA health: Ensure both HA members are synchronized before applying the change.
• It is not recommended to make this change from the FortiManager.

Related document:

Enhanced load balancing for LAG interfaces for NP6 platforms