Manual HA Failover Trigger Support (8.0.0)

FortiWeb 8.0.0 introduces the execute ha failover CLI command, enabling administrators to manually trigger a failover event in HA Active-Passive (AP), Active-Active Standard (AAS), or Active-Active High-Volume (AAH) environments. This command temporarily overrides the standard HA election logic to force the current node into standby mode, supporting controlled handover during maintenance, diagnostics, or testing scenarios.

Benefits

• Operational Flexibility – Enables manual promotion or demotion of HA members without requiring physical link changes or artificial fault injection.

• Troubleshooting Efficiency – Helps isolate issues by allowing role reversal in a controlled manner.

• Safe and Reversible – Failover status is automatically reset on reboot or when switching to Standalone mode.

Configuration

execute ha failover {set | unset | status}

set	Forces the current node to relinquish the primary role and enter standby.
unset	Clears the manual failover state and returns the node to standard HA election behavior.
status	Displays the current manual failover state of the node.

Operational Behavior

Scope and Availability

• Supported only in HA AP, AAS, or AAH modes.

• Not available in Standalone or Manager modes, including public cloud deployments.

Failover Control and Reset Conditions

• The manual failover state is automatically cleared under the following conditions:

  • The system reboots.

  • The HA mode is changed to Standalone.

HA Election Logic (when failover is cleared)

• The device resumes standard HA election based on:

  • Link status of monitor interfaces

  • HA uptime

  • HA priority

  • Serial number (used as a tie-breaker)

• If override is enabled, HA priority would take precedence over HA uptime.

Command Visibility and Logging

• Failover operations (set and unset) generate entries in the event log.

• The current manual failover state appears in the outputs of:

  • get system status

  • get system ha-status

Cluster Behavior

• If a new node joins a cluster where manual failover is enabled, the original HA election state is preserved.

• After full synchronization, member nodes retain their failover status unless explicitly reconfigured.