Fortinet FortiGate 4000 Betriebsanweisung

If you are configuring an active-active cluster, select a load balancing schedule.

Enable or disable HA heartbeat communication and set the heartbeat priority for each 
interface in the cluster.

By default HA heartbeat communication is set for two interfaces. You can disable HA 
heartbeat communication for either of these interfaces or enable HA heartbeat for 
other interfaces. In most cases you can maintain the default heartbeat device 
configuration as long as you can connect the heartbeat device interfaces together.

To enable HA heartbeat communication for an interface, enter a priority for the 
interface. To disable HA heartbeat communication for an interface, delete the priority 
for the interface.

The HA heartbeat priority range is 0 to 512. The interface with the highest priority 
handles all HA heartbeat traffic. If this interface fails or becomes disconnected, the 
interface with the next highest priority handles all HA heartbeat traffic.

The cluster units use the ethernet interfaces configured with HA heartbeat priorities for 
HA heartbeat communication. The HA heartbeat communicates cluster session 
information, synchronizes the cluster configuration, synchronizes the cluster routing 
table, and reports individual cluster member status. The HA heartbeat constantly 
communicates HA status information to make sure that the cluster is operating 
properly.

No load balancing. Select None when the cluster interfaces are connected 

to load balancing switches.

Load balancing if the cluster interfaces are connected to a hub. Traffic is 

distributed to cluster units based on the Source IP and Destination IP of the 

packet.

Least connection load balancing. If the cluster units are connected using 

switches, select Least Connection to distribute network traffic to the cluster 

unit currently processing the fewest connections.

Round robin load balancing. If the cluster units are connected using 

switches, select Round-Robin to distribute network traffic to the next 

available cluster unit.

Weighted round robin load balancing. Similar to round robin, but weighted 

values are assigned to each of the units in a cluster based on their capacity 

and on how many connections they are currently processing. For example, 

the primary unit should have a lower weighted value because it handles 

scheduling and forwards traffic. Weighted round robin distributes traffic more 

evenly because units that are not processing traffic will be more likely to 

receive new connections than units that are very busy.

Random load balancing. If the cluster units are connected using switches, 

select Random to randomly distribute traffic to cluster units.

Load balancing according to IP address. If the cluster units are connected 

using switches, select IP to distribute traffic to units in a cluster based on the 

Source IP and Destination IP of the packet.

Load balancing according to IP address and port. If the cluster units are 

connected using switches, select IP Port to distribute traffic to units in a 

cluster based on the source IP, source port, destination IP, and destination 

port of the packet.