【DB宝42】MySQL高可用架构MHA+ProxySQL实现读写分离和负载均衡（5）

华为云博客 技术博客 12753 28

六、故障切换

在Manager节点检查SSH、复制及MHA的状态。

docker exec -it MHA-LHR-Monitor-ip134 bash
masterha_check_ssh --conf=/etc/mha/mha.cnf
masterha_check_repl --conf=/etc/mha/mha.cnf
masterha_check_status --conf=/etc/mha/mha.cnf

-- 启动MHA监控进程
nohup masterha_manager --conf=/etc/mha/mha.cnf  --ignore_last_failover < /dev/null > /usr/local/mha/manager_start.log 2>&1 &

--关闭MHA监控进程
masterha_stop --conf=/etc/mha/mha.cnf


[root@MHA-LHR-Monitor-ip134 /]# masterha_check_status --conf=/etc/mha/mha.cnf
mha (pid:3738) is running(0:PING_OK), master:192.168.68.131

接下来，宕掉主库，继续观察ProxySQL的情况：

 -- 宕掉主库
 docker stop MHA-LHR-Master1-ip131

MHA自动执行了故障转移，主库切换为132，并发送告警邮件：

此时，来查看ProxySQL的情况：

MySQL [(none)]> select * from mysql_servers;
+--------------+----------------+------+-----------+--------+--------+-------------+-----------------+---------------------+---------+----------------+---------+
| hostgroup_id | hostname       | port | gtid_port | status | weight | compression | max_connections | max_replication_lag | use_ssl | max_latency_ms | comment |
+--------------+----------------+------+-----------+--------+--------+-------------+-----------------+---------------------+---------+----------------+---------+
| 10           | 192.168.68.132 | 3306 | 0         | ONLINE | 1      | 0           | 1000            | 0                   | 0       | 0              |         |
| 20           | 192.168.68.131 | 3306 | 0         | ONLINE | 1      | 0           | 1000            | 0                   | 0       | 0              |         |
| 20           | 192.168.68.133 | 3306 | 0         | ONLINE | 1      | 0           | 1000            | 0                   | 0       | 0              |         |
| 20           | 192.168.68.132 | 3306 | 0         | ONLINE | 1      | 0           | 1000            | 0                   | 0       | 0              |         |
+--------------+----------------+------+-----------+--------+--------+-------------+-----------------+---------------------+---------+----------------+---------+
4 rows in set (0.05 sec)

MySQL [(none)]> select * from runtime_mysql_servers;
+--------------+----------------+------+-----------+---------+--------+-------------+-----------------+---------------------+---------+----------------+---------+
| hostgroup_id | hostname       | port | gtid_port | status  | weight | compression | max_connections | max_replication_lag | use_ssl | max_latency_ms | comment |
+--------------+----------------+------+-----------+---------+--------+-------------+-----------------+---------------------+---------+----------------+---------+
| 10           | 192.168.68.132 | 3306 | 0         | ONLINE  | 1      | 0           | 1000            | 0                   | 0       | 0              |         |
| 20           | 192.168.68.131 | 3306 | 0         | SHUNNED | 1      | 0           | 1000            | 0                   | 0       | 0              |         |
| 20           | 192.168.68.133 | 3306 | 0         | ONLINE  | 1      | 0           | 1000            | 0                   | 0       | 0              |         |
| 20           | 192.168.68.132 | 3306 | 0         | ONLINE  | 1      | 0           | 1000            | 0                   | 0       | 0              |         |
+--------------+----------------+------+-----------+---------+--------+-------------+-----------------+---------------------+---------+----------------+---------+
4 rows in set (1.26 sec)

可以发现131已经变成SHUNNED状态，ProxySQL会避开这个主机。

此时再做压测等操作，所有负载会被分配到132和133上，此处不再测试。

接下来启动131，并以从库的身份加入原主从环境：

-- 启动131
docker start MHA-LHR-Master1-ip131

-- 在134的日志文件中找到恢复的语句
grep "All other slaves should start replication from here" /usr/local/mha/manager_running.log

-- 在131上执行恢复
mysql -uroot -plhr -h192.168.68.131 -P3306
CHANGE MASTER TO MASTER_HOST='192.168.68.132', 
MASTER_PORT=3306, 
MASTER_AUTO_POSITION=1,
MASTER_USER='repl',
MASTER_PASSWORD='lhr';

start slave;
show slave status \G

-- 设置只读
set global read_only=1;

查询ProxySQL：

MySQL [(none)]> select * from mysql_servers;
+--------------+----------------+------+-----------+--------+--------+-------------+-----------------+---------------------+---------+----------------+---------+
| hostgroup_id | hostname       | port | gtid_port | status | weight | compression | max_connections | max_replication_lag | use_ssl | max_latency_ms | comment |
+--------------+----------------+------+-----------+--------+--------+-------------+-----------------+---------------------+---------+----------------+---------+
| 10           | 192.168.68.132 | 3306 | 0         | ONLINE | 1      | 0           | 1000            | 0                   | 0       | 0              |         |
| 20           | 192.168.68.131 | 3306 | 0         | ONLINE | 1      | 0           | 1000            | 0                   | 0       | 0              |         |
| 20           | 192.168.68.133 | 3306 | 0         | ONLINE | 1      | 0           | 1000            | 0                   | 0       | 0              |         |
| 20           | 192.168.68.132 | 3306 | 0         | ONLINE | 1      | 0           | 1000            | 0                   | 0       | 0              |         |
+--------------+----------------+------+-----------+--------+--------+-------------+-----------------+---------------------+---------+----------------+---------+
4 rows in set (0.06 sec)

可以看到131为只读。若想让132只写，则可以删除相关记录：

MySQL [(none)]> delete from mysql_servers where hostgroup_id=20 and  hostname='192.168.68.132';
Query OK, 1 row affected (0.06 sec)

MySQL [(none)]> load mysql servers to runtime;
Query OK, 0 rows affected (0.68 sec)

MySQL [(none)]> save mysql servers to disk;
Query OK, 0 rows affected (0.10 sec)

MySQL [(none)]> select * from mysql_servers;
+--------------+----------------+------+-----------+--------+--------+-------------+-----------------+---------------------+---------+----------------+---------+
| hostgroup_id | hostname       | port | gtid_port | status | weight | compression | max_connections | max_replication_lag | use_ssl | max_latency_ms | comment |
+--------------+----------------+------+-----------+--------+--------+-------------+-----------------+---------------------+---------+----------------+---------+
| 10           | 192.168.68.132 | 3306 | 0         | ONLINE | 1      | 0           | 1000            | 0                   | 0       | 0              |         |
| 20           | 192.168.68.131 | 3306 | 0         | ONLINE | 1      | 0           | 1000            | 0                   | 0       | 0              |         |
| 20           | 192.168.68.133 | 3306 | 0         | ONLINE | 1      | 0           | 1000            | 0                   | 0       | 0              |         |
+--------------+----------------+------+-----------+--------+--------+-------------+-----------------+---------------------+---------+----------------+---------+
3 rows in set (0.05 sec)

MySQL [(none)]> select * from runtime_mysql_servers;
+--------------+----------------+------+-----------+--------+--------+-------------+-----------------+---------------------+---------+----------------+---------+
| hostgroup_id | hostname       | port | gtid_port | status | weight | compression | max_connections | max_replication_lag | use_ssl | max_latency_ms | comment |
+--------------+----------------+------+-----------+--------+--------+-------------+-----------------+---------------------+---------+----------------+---------+
| 10           | 192.168.68.132 | 3306 | 0         | ONLINE | 1      | 0           | 1000            | 0                   | 0       | 0              |         |
| 20           | 192.168.68.133 | 3306 | 0         | ONLINE | 1      | 0           | 1000            | 0                   | 0       | 0              |         |
| 20           | 192.168.68.131 | 3306 | 0         | ONLINE | 1      | 0           | 1000            | 0                   | 0       | 0              |         |
+--------------+----------------+------+-----------+--------+--------+-------------+-----------------+---------------------+---------+----------------+---------+
3 rows in set (0.94 sec)

可以看到132为主库，131和133为从库。Orchestrator界面：

七、界面监控结果（ProxySQL本身+PMM监控）

7.1 ProxySQL自带监控

ProxySQL监控结果：

7.2 PMM监控

监控可以使用ProxySQL本身自动的监控，也可以使用PMM来监控。PMM（Percona Monitoring and Management）是一个免费的开源平台，用于管理和监视数据库性能，在docker环境中即可运行。它可以主动管理和监控MySQL（AWS RDS MySQL、Aurora MySQL、用户自建MySQL实例）、MariaDB、MongoDB、PostgreSQL等数据库，也可以监控ProxySQL中间件，并提供了众多指标与多样告警方式。

PMM监控ProxySQL的命令如下：

pmm-admin add proxysql --username=root --password=lhr  --host=192.168.66.35 --port=26032 --service-name=proxysql-192.168.66.35-26032

PMM监控结果：


About Me

● 本文作者：小麦苗，部分内容整理自网络，若有侵权请联系小麦苗删除
● 本文在个人微 信公众号（DB宝）上有同步更新
● QQ群号： 230161599 、618766405，微信群私聊
● 个人QQ号（646634621），微 信号（db_bao），注明添加缘由
● 于 2021年3月 在西安完成
● 最新修改时间：2021年3月
● 版权所有，欢迎分享本文，转载请保留出处

●小麦苗的微店： https://weidian.com/?userid=793741433
●小麦苗出版的数据库类丛书： http://blog.itpub.net/26736162/viewspace-2142121/
●小麦苗OCP、OCM、高可用、DBA学习班（Oracle、MySQL、NoSQL）： http://blog.itpub.net/26736162/viewspace-2148098/
●数据库笔试面试题库及解答： https://mp.weixin.qq.com/s/Vm5PqNcDcITkOr9cQg6T7w

使用微信客户端扫描下面的二维码来关注小麦苗的微信公众号（DB宝）及QQ群（DBA宝典）、添加小麦苗微信， 学习最实用的数据库技术。
小麦苗信息

（完）