Document how to run database replication from the last data checkpoint
I think we may need to create one more extra replication slot when doing a planned failover, this will allow us to not replicate database data from scratch when we bring the primary back online.