Yet another feature has been included in Amazon RDS for MySQL. It now supports for MySQL’s “Promote Read Replica” feature, which means you can take a database replica and convert it to a standalone database.
Promote Read Replica support has advantages in several situations, including database recovery. While Amazon RDS has its own set of data recovery features, including Multi-AZ deployments and Point in Time Recovery, having another option is still useful.
The main disadvantage of using a Read Replica as a backup database if other options fail is that with asynchronous replication, database writes occur on a Read Replica after they have already occurred on the Source DB Instance. If the workload is heavy, this replication can lag significantly behind, so the data recovered would not include the most recent updates.
Another suggested use of Promote Read Replica is to give you a way to carry out operations that take a long time such as creating and re-building indexes. Such operations can be tough on the performance of your DB Instance, so the suggestion is that you perform the operations on a Read Replica, and once the operations are complete and the updates are caught up with the Source DB Instance, you can promote the Read Replica, and point your applications to it.
A final way Promote Read Replicas can be useful is when sharding. When a large database is split into smaller databases, you can split tables that don’t share queries onto different hosts, or duplicate tables across multiple hosts then use a hashing algorithm to work out which host to use for a particular row. You can create a Read Replica for each shard, then promote them when you decide to convert them into standalone shards. Any rows or tables belonging to the other shards can then be deleted.