The Buffalo TeraStation™ supports fail over fault tolerance in the network.
Fail Over is different to Replication. In Fail Over scenario there are at least two TeraStation present in one network but only one TeraStation can be seen as a device. The second unit is hidden but directly communicating with the first unit. So the data stored on the first is mirrored to the second immediately. The “magic” behind is the behavior of the second in case of fault of the first. As soon as the first unit disappears from the network – maybe due to power failure, network failure or whatever – the second, hidden unit appears and takes over the place from the first one. The failover mechanism allows the second unit to act identically like the first, unavailable unit. Even the Mac-Addresses, the hostname, the IP-Addresses, the complete “personality” of the first one is taken by the second one.
Failover means that in case of unavailability there is a little gap in time until everything works again from the perspective of the user. Behind the scene the TeraStation exchanged their positions to avoid any lack in workflow and operation.
Failover is based on the Spanning Tree control mechanism.