Yesterday a power surge took out our SQL Server. Had to have the motherboard replaced and everything. Not good.
Anyway, I'm now looking into failover clustering to prevent such an event from impacting our business.
Is the concept of failover clustering as follows?
Obtain a new server to act as a node
Install SQL Server on this server, set up as a failover node
Changes to the "live" database are replicated on the failover node
If the "live" server goes down, the failover node assumes the identity of the main SQL Server. For example, if the "live" server is named "CO-SQL", the node will then assume that identity so it can be found on the network.
Is that about right?