The backend provides a redundancy switching mechanism for M&C systems connected to it. A redundant pair of M&C systems share the M&C name, but the M&C systems have different IP addresses. The redundancy works asymmetrically (there is one primary and one backup M&C) and one-way (redundancy switch will never happen from backup to primary).
If the communication to the primary M&C breaks and cannot be re-established for some time, the backend stops the M&C service on the primary machine and starts it on the backup machine. From now on the backend connects to the IP of backup M&C to control it.
All configuration about this is done in the 'backend.properties' file. The second IP address for the backup M&C and parameters how long the backend shall retry to connect to the primary M&C before it does the switch over to the backup M&C.
The actual state of the redundancy for each M&C (pair) is reported through the /api/v1/mnclist API call. Each MncInfo record in the returned list defines a 'redundancy' field which tells about the actual state.