FortiGate-6000 and 7000 interface status LEDs do not accurately show link status for copper SPF transceivers that use the RX_LOS pin to indicate link status.

In some cases on a FortiGate-7000, configuration changes may not successfully be synchronized to all of the FPMs. If this occurs you can run the command killall confsyncd on the FPM that is out of sync to stop the confsyncd process. The process should restart and re-synchronize the configuration; which can take about 30 seconds.

After an FGCP HA failover, its possible that traffic may not be able to pass through a transparent mode VDOM until the FDB MAC table expires in approximately 6 minutes.

If you factory reset the backup FortiGate-6000 or 7000 in an FGCP cluster and then re-configure the backup FortiGate to join with the correct HA settings, the backup FortiGate will not be able to join the cluster again until it is manually restarted.

In an FGSP configuration with asymmetric routing and four FortiGate-6000s or 7000s, the first reply packet of all new sessions is dropped.

When a FortiGate-6000 or 7000 is idle, the src-vis process can use excessive amounts of CPU resources and cause 100% softirq usage on each CPU on the primary FPC or FPM.

A graceful upgrade of an FGCP HA cluster may fail if the configuration includes a large number of VDOMs. The upgrade process times out waiting for the backup FortiGate-6000 or 7000 to upgrade and start up.