I don’t know about best practice, but I’d host the db backend outside of the cluster. You could create another cluster and host the db there. That plus an external backup of the cluster db would get you a fault tolerance and disaster recovery .
To recap , here is my idea
Create 2 cluster running on different hardware.
Put cluster 1’s external db in cluster 2
Put cluster 2’s exernal db in cluster 1
Use a cron job or similar to back up both db’s into a NAS or other storage.