Documentation Home
MySQL 8.0 Reference Manual
Related Documentation Download this Manual
PDF (US Ltr) - 40.3Mb
PDF (A4) - 40.4Mb
PDF (RPM) - 39.9Mb
HTML Download (TGZ) - 10.6Mb
HTML Download (Zip) - 10.6Mb
HTML Download (RPM) - 9.2Mb
Man Pages (TGZ) - 241.2Kb
Man Pages (Zip) - 346.4Kb
Info (Gzip) - 3.9Mb
Info (Zip) - 3.9Mb
Excerpts from this Manual

MySQL 8.0 Reference Manual  /  ...  /  Deploying InnoDB ReplicaSet

21.3.2 Deploying InnoDB ReplicaSet

You deploy InnoDB ReplicaSet in a similar way to InnoDB Cluster. First you configure some MySQL server instances, the minimum is two instances, see Section 21.1, “MySQL AdminAPI”. One functions as the primary, in this tutorial rs-1; the other instance functions as the secondary, in this tutorial rs-2; which replicates the transactions applied by the primary. This is the equivalent of the source and replica known from asynchronous MySQL replication. Then you connect to one of the instances using MySQL Shell, and create a ReplicaSet. Once the ReplicaSet has been created, you can add instances to it.

InnoDB ReplicaSet is compatible with sandbox instances, which you can use to deploy locally, for example for testing purposes. See Deploying Sandbox Instances for instructions. However, this tutorial assumes you are deploying a production InnoDB ReplicaSet, where each instance is running on a different host.

InnoDB ReplicaSet Prerequisites

To use InnoDB ReplicaSet you should be aware of the following prerequisites:

  • Only instances running MySQL version 8.0 and later are supported

  • Only GTID-based replication is supported, binary log file position replication is not compatible with InnoDB ReplicaSet

  • Only Row Based Replication (RBR) is supported, Statement Based Replication (SBR) is unsupported

  • Replication filters are not supported

  • Unmanaged replication channels are not allowed on any instance

  • A ReplicaSet consists of maximum one primary instance, and one or multiple secondaries are supported. Although there is no limit to the number of secondaries you can add to a ReplicaSet, each MySQL Router connected to a ReplicaSet has to monitor each instance. Therefore, the more instances that are added to a ReplicaSet, the more monitoring has to be done.

  • The ReplicaSet must be entirely managed by MySQL Shell. For example, the replication account is created and managed by MySQL Shell. Making configuration changes to the instance outside of MySQL Shell, for example using SQL statements directly to change the primary, is not supported. Always use MySQL Shell to work with InnoDB ReplicaSet.

AdminAPI and InnoDB ReplicaSet enable you to work with MySQL replication without a deep understanding of the underlying concepts. However, for background information see Chapter 17, Replication.

Configuring InnoDB ReplicaSet Instances

Use dba.configureReplicaSetInstance(instance) to configure each instance you want to use in your replica set. MySQL Shell can either connect to an instance and then configure it, or you can pass in instance to configure a specific remote instance. To use an instance in a ReplicaSet, it must support persisting settings. See Persisting Settings.

When you connect to the instance for administration tasks you require a user with suitable privileges. The preferred method to create users to administer a ReplicaSet is using the setupAdminAccount() operation. See Creating User Accounts for Administration. Alternatively, the dba.configureReplicaSetInstance() operation can optionally create an administrator account, if the clusterAdmin option is provided. The account is created with the correct set of privileges required to manage InnoDB ReplicaSet.

Tip

The administrator account must have the same user name and password across all instances of the same cluster or replica set.

To configure the instance at rs-1:3306, with a cluster administrator named rsadmin issue:

mysql-js> dba.configureReplicaSetInstance('root@rs-1:3306', {clusterAdmin: "'rsadmin'@'rs-1%'"});

The interactive prompt requests the password required by the specified user. To configure the instance MySQL Shell is currently connected to, you can specify a null instance definition. For example issue:

mysql-js> dba.configureReplicaSetInstance('', {clusterAdmin: "'rsadmin'@'rs-1%'"});

The interactive prompt requests the password required by the specified user. This checks the instance which MySQL Shell is currently connected to is valid for use in an InnoDB ReplicaSet. Settings which are not compatible with InnoDB ReplicaSet are configured if possible. The cluster administrator account is created with the privileges required for InnoDB ReplicaSet.

Creating an InnoDB ReplicaSet

Once you have configured your instances, connect to an instance and use dba.createReplicaSet() to create a managed ReplicaSet that uses MySQL asynchronous replication, as opposed to MySQL Group Replication used by InnoDB Cluster. The MySQL instance which MySQL Shell is currently connected to is used as the initial primary of the ReplicaSet.

The dba.createReplicaSet() operation performs several checks to ensure that the instance state and configuration are compatible with a managed ReplicaSet and if so, a metadata schema is initialized on the instance. If you want to check the operation but not actually make any changes to the instances, use the dryRun option. This checks and shows what actions the MySQL Shell would take to create the ReplicaSet. If the ReplicaSet is created successfully, a ReplicaSet object is returned. Therefore it is best practice to assign the returned ReplicaSet to a variable. This enables you to work with the ReplicaSet, for example by calling the ReplicaSet.status() operation. To create a ReplicaSet named example on instance rs-1 and assign it to the rs variable, issue:

mysql-js> \connect root@rs-1:3306
...
mysql-js> var rs = dba.createReplicaSet("example")
A new replicaset with instance 'rs-1:3306' will be created.

* Checking MySQL instance at rs-1:3306

This instance reports its own address as rs-1:3306
rs-1:3306: Instance configuration is suitable.

* Updating metadata...

ReplicaSet object successfully created for rs-1:3306.
Use rs.add_instance() to add more asynchronously replicated instances to this replicaset
and rs.status() to check its status.

To verify that the operation was successful, you work with the returned ReplicaSet object. For example this provides the ReplicaSet.status() operation, which displays information about the ReplicaSet. We already assigned the returned ReplicaSet to the variable rs, so issue:

mysql-js> rs.status()
{
    "replicaSet": {
        "name": "example", 
        "primary": "rs-1:3306", 
        "status": "AVAILABLE", 
        "statusText": "All instances available.", 
        "topology": {
            "rs-1:3306": {
                "address": "rs-1:3306", 
                "instanceRole": "PRIMARY", 
                "mode": "R/W", 
                "status": "ONLINE"
            }
        }, 
        "type": "ASYNC"
    }
}

This output shows that the ReplicaSet named example has been created, and that the primary is rs-1. Currently there is only one instance, and the next task is to add more instances to the ReplicaSet.