You are here: Product Documentation > Unified Threat Management > About the System Menu > System > Hardware > Failover

Managing Hardware Failover

Note: This topic applies to the Hearst Release.

Note: A Unified Threat Management serial is required to configure a failover installation. For more information, contact your Smoothwall representative.

The Smoothwall’s hardware failover enables you to configure a failover which, in the event of hardware failure, provides all the protection and services your primary Smoothwall usually provides.

When configured and enabled, the failover Smoothwall runs in a standby mode monitoring the primary Smoothwall for a heartbeat communication. Heartbeat is the name of a suite of services and configuration options that enable two identical Smoothwalls to be configured to provide hardware failover.

The primary system periodically copies settings to the failover unit to ensure that the failover unit can provide a fully configured service if the primary system fails.

Note: Settings are copied intermittently and it is theoretically possible that the failover unit will be a few minutes behind configuration changes made to the primary system.

If the primary Smoothwall fails, it stops responding to the failover unit’s heartbeat and the failover unit therefore determines that the primary system is no longer available. This occurs somewhere between 0 seconds and the Keep-alive internal time specified when configuring failover.

The failover unit then enters a more responsive mode where it monitors the primary system for its revival. It remains in this mode for the length of Dead time you have configured. This stage is designed principally to cope with intermittent failures within the communication system, such as a heavily loaded primary system.

Once Dead time has expired, the failover unit awakens from its standby mode and begins re-instating the settings and services allowing it to take over operations from the primary system. Since part of this information includes the IP addresses for each of the primary system’s interfaces, the failover unit essentially provides a drop-in replacement and the transition generally go unnoticed.

When the primary system starts to respond again, be it minutes, days or weeks later, assuming that Auto failback is enabled, the failover unit hands over control to the master, de-activates its configuration and services, and returns to standby mode.