When designing the solution, it is recommended that IPStor servers be configured as a fail-over pair.  In this configuration, there will be no single point of failure.  The IPStor servers communicate using a heartbeat process. Should there be a hardware failure on one server, the other member of the pair will take up all I/O processing. Upon restarting the failed server, service is automatically switched back when the machine comes back online.  Administrative forced failovers are also possible, a useful feature when server maintenance is required. One server can temporarily handle a dual traffic load while the second server is serviced (e.g. additional memory is added, new storage is attached, etc.).

All connections between client and storage can also be configured in pairs, providing a fully redundant solution from end-to-end.