Shutdown the cluster.
Restore the Publish node using Restoring the Backup.
Start the master node followed by the slave node of the author cluster.