Since it is quite common that there is only one specific free space per device, the file storage should be able to work with several mounted devices.
The following assumptions can be made:
all devices are available under one root point. For example /opt/data/0001, /opt/data/0002...
when writing, a device is selected at random. Devices that are less busy can experience a higher probability of allocation.
busy devices are no longer taken into account.
a file in the root of the device can contain information about the decision. (Size, occupancy, free space...)
In the event of a write error, the device can be automatically switched to read-only mode. The operation then continues on another device
Since it is quite common that there is only one specific free space per device, the file storage should be able to work with several mounted devices. The following assumptions can be made: