The failover feature allows back ends to automatically switch to a different server if the primary server fails. The list of servers is given as a comma-separated list; any number of spaces is allowed around the comma. The servers are listed in order of preference. The list can contain any number of servers. The failover mechanism distinguishes between a machine and a service. The back end first tries to resolve the hostname of a given machine; if this resolution attempt fails, the machine is considered offline. No further attempts are made to connect to this machine for any other service. If the resolution attempt succeeds, the back end tries to connect to a service on this machine. If the service connection attempt fails, then only this particular service is considered offline and the back end automatically switches over to the next service. The machine is still considered online and might still be tried for another service. Further connection attempts are made to machines or services marked as offline after a specified period of time; this is currently hard coded to 30 seconds. If there are no more machines to try, the back end as a whole switches to offline mode, and then attempts to reconnect every 30 seconds.