实验环境

主机名/IP	角色
server01/192.168.1.201	主
server02/192.168.1.202	辅
server03/192.168.1.203	辅
server04/192.168.1.204	只读副本

InnoDB Cluster 8.4部署步骤请查阅文档：https://www.wakamizu.cn/innodb-cluster-8-4%e9%9b%86%e7%be%a4%e9%83%a8%e7%bd%b2/

server04

添加映射

vim /etc/hosts

192.168.1.201 server01
192.168.1.202 server02
192.168.1.203 server03
192.168.1.204 server04

安装mysql存储库

wget https://dev.mysql.com/get/mysql-apt-config_0.8.32-1_all.deb
dpkg -i mysql-apt-config_0.8.32-1_all.deb 
apt update -y

安装mysql

apt install mysql-server -y

配置允许外部访问mysql

vim /etc/mysql/mysql.conf.d/mysql.conf

[mysqld]
bind-address=0.0.0.0

重启服务

systemctl restart mysql

安装mysql shell

apt install mysql-shell -y

使用mysql shell连接mysql

mysqlsh mysql://root@localhost:3306

进入js模式查询已连接数据库信息

\js
session;
shell.status();
\quit

执行mysql指令可以使用：

\sql mysql指令

创建集群服务配置账户

mysqlsh mysql://root@localhost:3306
\js
dba.configureInstance('root@localhost:3306',{clusterAdmin:"'clusteradmin'@'%'"});

检查实例是否符合集群要求

dba.checkInstanceConfiguration('clusteradmin@localhost:3306');

server01

将server04加入集群

mysqlsh -u root -p
\js
var cluster = dba.getCluster()
cluster.addReplicaInstance('clusteradmin@192.168.1.204:3306')

查看信息

cluster.status()

{
    "clusterName": "testCluster", 
    "defaultReplicaSet": {
        "name": "default", 
        "primary": "server01:3306", 
        "ssl": "REQUIRED", 
        "status": "OK", 
        "statusText": "Cluster is ONLINE and can tolerate up to ONE failure.", 
        "topology": {
            "server01:3306": {
                "address": "server01:3306", 
                "memberRole": "PRIMARY", 
                "mode": "R/W", 
                "readReplicas": {
                    "Replica1": {
                        "address": "server04:3306", 
                        "role": "READ_REPLICA", 
                        "status": "ONLINE", 
                        "version": "8.4.2"
                    }
                }, 
                "replicationLag": "applier_queue_applied", 
                "role": "HA", 
                "status": "ONLINE", 
                "version": "8.4.2"
            }, 
            "server02:3306": {
                "address": "server02:3306", 
                "memberRole": "SECONDARY", 
                "mode": "R/O", 
                "readReplicas": {}, 
                "replicationLag": "applier_queue_applied", 
                "role": "HA", 
                "status": "ONLINE", 
                "version": "8.4.2"
            }, 
            "server03:3306": {
                "address": "server03:3306", 
                "memberRole": "SECONDARY", 
                "mode": "R/O", 
                "readReplicas": {}, 
                "replicationLag": "applier_queue_applied", 
                "role": "HA", 
                "status": "ONLINE", 
                "version": "8.4.2"
            }
        }, 
        "topologyMode": "Single-Primary"
    }, 
    "groupInformationSourceMember": "server01:3306"
}

加入集群指令也可以增加参数指定克隆来源

cluster.addReplicaInstance('clusteradmin@192.168.1.204:3306',{ recoveryMethod: "clone",cloneDonor:"192.168.1.202:3306"})

注意：当Group Replication以单主模式运行时，无法在次要成员上启动复制IO线程。这是因为在单主模式下，只有主节点可以执行写操作，而次要节点只能执行读操作。

移除只读副本节点

cluster.removeInstance('clusteradmin@192.168.1.204:3306')

重新加入集群

cluster.rejoinInstance('clusteradmin@192.168.1.204:3306')