MySQL :: MySQL Shell 8.3 :: 7.11 MySQL InnoDB Cluster Read Replicas

8.4

MySQL Shell 8.3 / MySQL InnoDB Cluster / MySQL InnoDB Cluster Read Replicas

7.11 MySQL InnoDB Cluster Read Replicas

7.11.1 Prerequisites
7.11.2 Creating Read Replicas
7.11.3 Modifying or Removing Read Replicas
7.11.4 Monitoring Read Replicas

MySQL Shell Read Replicas are read-only copies of a Cluster member. Asynchronous replication keeps the replica up to date, enabling you to scale out your workload, offload read requests from your Cluster to one, or more, dedicated read-only instances, and provide additional redundancy to your dataset. In the event of a failure of the source, the Read Replica automatically connects to another Cluster member and resumes replication.

Figure 7.3 InnoDB Cluster Read Replicas overview

Three MySQL servers are grouped together as a high availability Cluster. One of the servers is the read/write primary instance, and the other two are read-only secondary instances. Group Replication is used to replicate data from the primary instance to the secondary instances. A group of Read Replicas replicates from the Cluster's primary by asynchronous replication. MySQL Router connects client applications (in this example, a MySQL Connector) to the primary instance with a read-write connection and to the Read Replicas with a read-only connection.

Read Replicas can be configured to replicate from any member of the Cluster. By default, they replicate from the primary. You can also promote Read Replicas to secondary Cluster members and demote secondary Cluster members to Read Replicas.

PREV HOME UP NEXT