The cluster service must be stopped to modify the shares, and NFS settings. Due to the nature of the 'failover-data' share, it can not be dynamically changed, as each server needs to be configured, and the service restarted from the primary node.