In a FortiManager High Availability (HA) cluster, the primary unit's configuration defines the cluster membership. When a secondary (peer) device fails and cannot be recovered, the HA cluster is in a degraded state. To return the cluster to a stable, working condition with the remaining active members, an administrator must manually intervene. The correct procedure is to access the primary FortiManager's HA settings and remove the configuration entry, including the IP address, for the failed secondary device. This action updates the cluster's topology and resolves the inconsistency caused by the unreachable member.

A: This is incorrect. Manual mode HA requires administrator intervention to manage cluster membership, especially in a failure scenario. The process is not transparent.

B: Rebooting the failed device will not change the HA configuration on the primary unit. If the device is permanently offline, this action has no effect on the cluster's state.

C: Promoting a secondary device is the procedure for a primary device failure, not a secondary one. The primary unit is still operational in this scenario.

1. FortiManager Administration Guide 7.4.1

Chapter: "High Availability"

Section: "Managing the cluster"

Subsection: "Removing a failed cluster member"

Page 71.

The documentation explicitly states: "If a cluster member fails

you must remove it from the cluster. You must edit the HA configuration on the primary FortiManager to remove the failed device." It then provides the GUI steps: "Go to System Settings > HA... In the Peer Information table

select the failed device and click Delete." This directly supports the action of reconfiguring the primary device.