Configure FireCluster Advanced Settings
In Policy Manager, the Advanced tab in the FireCluster Configuration dialog box includes settings for logging and notification, and enables you to adjust the lost heartbeat threshold.
Configure Logging and Notification
Log messages are always created for FireCluster events.
To configure notification settings for FireCluster failover and failback events:
- Click Notification.
- Select a notification method: SNMP trap, email message, or pop-up window.
For more information about notification settings, see Set Logging and Notification Preferences.
To set the diagnostic log level for FireCluster events, from Policy Manager:
- Select Setup > Logging.
- Click Diagnostic Log Level.
For more information about diagnostic logging, see Set the Diagnostic Log Level.
Change the Lost Heartbeat Threshold
The cluster master sends a VRRP heartbeat packet through the primary and backup cluster interfaces once per second. The Lost Heartbeat Threshold determines the number of consecutive heartbeats not received by the backup master before a FireCluster failover is triggered.
The default Lost Heartbeat Threshold value of three is optimal for most FireCluster configurations. Before you change the default value, make sure you eliminate any legitimate causes for failover. Look in the log messages to see if the lost heartbeat was caused by something else that happened. To see the cause of lost heartbeats, you might need to temporarily increase the diagnostic log level, as described in the previous section.
If unexplained failovers occur with your FireCluster, with no known cause, you can increase the Lost Heartbeat Threshold to try to make the cluster more stable. The maximum value for Lost Heartbeat Threshold is 10.
To change the lost heartbeat threshold:
In the Lost Heartbeat Threshold text box, type or select a value between 3 and 10.
Use Hardware Status as a Criteria for FireCluster Failover
FireCluster failover can be triggered based upon a comparison of the weighted average index (WAI) of each cluster member. By default the hardware health index (HHI) is not used in the calculation of the WAI.
To enable the HHI to be used in the calculation of the weighted average index:
Select the Monitor hardware status as a criteria for FireCluster failover check box.
For more information about the hardware health index and the calculation of WAI, see Monitor Cluster Health.