MySQL高可用方案
投票选举机制，较复杂
MySQL本身没有提供replication failover的解决方案，自动切换需要依赖MHA脚本
可以有多台从库，从库可以做报表和备份
MySQL复制技术
===========================================================================
重置数据库：
# service mysqld stop
# rm -rf /usr/local/mysql/data/*
# /usr/local/mysql/scripts/mysql_install_db --user=mysql --basedir=/usr/local/mysql --datadir=/usr/local/mysql/data
复制拓扑：
复制原理：
1. 在主库上把数据更改记录到二进制日志（Binary Log）中。
2. 备库将主库上的日志复制到自己的中继日志（Relay Log）中。
3. 备库读取中继日志中的事件，将其重放到备库数据库之上。
一、主/备均为刚初始的数据库
单主到多备： Master-MultiSlave
Master Slave1 Slave2
＝＝＝＝＝＝＝＝＝＝＝＝＝＝＝＝＝＝＝＝＝＝＝＝＝＝＝＝＝＝＝＝
IP 192.168.1.27 192.168.1.66 192.168.1.251
Server_ID 27 66 251
如何实现主从复制？
在主服务器（master）上
启用二进制日志
选择一个唯一的server-id
创建具有复制权限的用户
在从服务器（slave）上
启用中继日志（二进制日志可开启，也可不开启）
选择一个唯一的server-id
连接至主服务器，并开始复制
1. 主库
[root@master ~]# vim /etc/my.cnf
[mysqld]
log-bin=master-bin
binlog_format = row
sync_binlog = 1
skip_name_resolv = 1
log_slave_updates = 1
server_id = 27
[root@master ~]# service mysqld start
[root@master ~]# mysql
mysql> reset master;
mysql> grant replication slave, replication client on *.*
-> to rep@'192.168.1.%' identified by 'localhost';
mysql> flush privileges;
2. 备库
a. 测试复制账号
[root@slave1 ~]# mysql -h 192.168.1.27 -urep -plocalhost
b. 配置复制
[root@slave1 ~]# vim /etc/my.cnf
log-bin=slave1-bin
binlog_format = row
sync_binlog = 1
skip_name_resolv = 1
log_slave_updates = 1
server_id = 66
[root@slave1 ~]# service mysqld start
[root@slave1 ~]# mysql
mysql> reset master;
mysql> change master to
-> master_host='192.168.1.27',
-> master_user='rep',
-> master_password='localhost',
-> master_log_file='master-bin.000001',
-> master_log_pos=0;
Query OK, 0 rows affected (0.02 sec)
mysql> show slave statusG
*************************** 1. row ***************************
Slave_IO_State:
Master_Host: 192.168.1.27
Master_User: rep
Master_Port: 3306
Connect_Retry: 60
Master_Log_File: mysql-bin.000001
Read_Master_Log_Pos: 4
Relay_Log_File: mysql-relay-bin.000001
Relay_Log_Pos: 4
Relay_Master_Log_File: mysql-bin.000001
Slave_IO_Running: No
Slave_SQL_Running: No
mysql> start slave;
Query OK, 0 rows affected (0.00 sec)
mysql> show slave statusG
*************************** 1. row ***************************
Slave_IO_State: Waiting for master to send event
Master_Host: 192.168.1.27
Master_User: rep
Master_Port: 3306
Connect_Retry: 60
Master_Log_File: mysql-bin.000001
Read_Master_Log_Pos: 354
Relay_Log_File: mysql-relay-bin.000002
Relay_Log_Pos: 500
Relay_Master_Log_File: mysql-bin.000001
Slave_IO_Running: Yes
Slave_SQL_Running: Yes
3. 测试
Master：
mysql> show processlistG
*************************** 2. row ***************************
Id: 2
User: rep
Host: 192.168.10.37:50915
db: NULL
Command: Binlog Dump
Time: 324
State: Master has sent all binlog to slave; waiting for binlog to be updated
Info: NULL
2 rows in set (0.00 sec)
mysql> create database bbs;
Query OK, 1 row affected (0.00 sec)
mysql> create table bbs.t1(id int);
Query OK, 0 rows affected (0.03 sec)
mysql> insert into bbs.t1 values(1);
Query OK, 1 row affected (0.02 sec)
mysql> select * from bbs.t1;
+------+
| id |
+------+
| 1 |
+------+
1 row in set (0.00 sec)
Slave：
mysql> show processlistG
*************************** 2. row ***************************
Id: 2
User: system user
Host:
db: NULL
Command: Connect
Time: 356
State: Waiting for master to send event
Info: NULL
*************************** 3. row ***************************
Id: 3
User: system user
Host:
db: NULL
Command: Connect
Time: -173772
State: Slave has read all relay log; waiting for the slave I/O thread to update it
Info: NULL
3 rows in set (0.00 sec)
mysql> select * from bbs.t1;
+------+
| id |
+------+
| 1 |
+------+
1 row in set (0.03 sec)
二、针对已经运行一段时间的主库实现主/备
单主到多备： Master-MultiSlave
Master Slave1 Slave2
＝＝＝＝＝＝＝＝＝＝＝＝＝＝＝＝＝＝＝＝＝＝＝＝＝＝＝＝＝＝＝＝
IP 192.168.1.27 192.168.1.66 192.168.1.251
Server_ID 27 66 251
1. 主库
[root@master ~]# vim /etc/my.cnf
log-bin=master-bin
binlog_format = row
sync_binlog = 1
skip_name_resolv = 1
log_slave_updates = 1
server_id = 27
[root@master ~]# service mysqld restart
[root@master ~]# mysql
mysql> grant replication slave, replication client on *.*
-> to rep@'192.168.1.%' identified by 'localhost';
mysql> flush privileges;
========================================================
初始化备库（使其和主库数据一致）： 逻辑备份，物理备份
主库：
mysql> flush tables with read lock; //主服务器锁定表
[root@master ~]# mysqldump --all-databases > all.sql
[root@master ~]# mysql -e 'show master status'
+------------------+----------+--------------+------------------+
| File | Position | Binlog_Do_DB | Binlog_Ignore_DB |
+------------------+----------+--------------+------------------+
| master-bin.000001 | 699 | | |
+------------------+----------+--------------+------------------+
mysql> unlock tables; //解锁表
[root@master ~]# rsync -va all.sql 192.168.1.66:/
========================================================
2. 备库
[root@slave1 ~]# vim /etc/my.cnf
server-id = 66
[root@slave1 ~]# service mysqld start
[root@slave1 ~]# mysql
mysql> reset master;
mysql> source /all.sql
mysql> change master to
master_host='192.168.1.27',
master_user='rep',
master_password='localhost',
master_log_file='master-bin.000001',
master_log_pos=699;
Query OK, 0 rows affected (0.02 sec)
mysql> start slave;
Query OK, 0 rows affected (0.00 sec)
mysql> show slave statusG
*************************** 1. row ***************************
Slave_IO_State: Waiting for master to send event
Master_Host: 192.168.1.27
Master_User: rep
Master_Port: 3306
Connect_Retry: 60
Master_Log_File: master-bin.000001
Read_Master_Log_Pos: 354
Relay_Log_File: mysql-relay-bin.000002
Relay_Log_Pos: 500
Relay_Master_Log_File: mysql-bin.000001
Slave_IO_Running: Yes
Slave_SQL_Running: Yes
MySQL主从复制的状况监测
主从状况监测主要参数
Slave_IO_Running: IO线程是否打开 YES/No/NULL
Slave_SQL_Running: SQL线程是否打开 YES/No/NULL
Seconds_Behind_Master: NULL #和主库比同步的延迟的秒数
可能导致主从延时的因素
主从时钟是否一致
网络通信是否存在延迟
是否和日志类型，数据过大有关
从库性能，有没开启binlog
从库查询是否优化
常见状态错误排除
发现IO进程错误，检查日志，排除故障：
# tail localhost.localdomain.err
...2015-11-18 10:55:50 3566 [ERROR] Slave I/O: Fatal error: The slave I/O thread stops because master and slave have equal MySQL server UUIDs; these UUIDs must be different for replication to work. Error_code: 1593
找到原因：从5.6开始复制引入了uuid的概念，各个复制结构中的server_uuid得保证不一样
解决方法：(从库是克隆机器)修改从库的uuid
# vim auto.cnf server-uuid=
show slave status;报错：Error xxx doesn’t exist
解决方法：
stop slave;
set global sql_slave_skip_counter = 1;
start slave;
三、常见复制拓朴
1. 一主库多备库
2. 主库，分发主库以及备库
3. 主——主复制（双主）
四、MySQL 主主同步
重置数据库：
# service mysqld stop
# rm -rf /usr/local/mysql/data/*
# /usr/local/mysql/scripts/mysql_install_db --user=mysql --basedir=/usr/local/mysql --datadir=/usr/local/mysql/data
1. mysql1 192.168.1.4:
[root@mysql1 ~]# vim /etc/my.cnf
log-bin=mysql-bin
server-id = 4
[root@mysql1 ~]# service mysqld start
[root@mysql1 ~]# mysql
mysql> reset master;
mysql> grant replication slave, replication client on *.*
-> to rep@'192.168.1.%' identified by 'localhost';
mysql> flush privileges;
mysql> change master to
-> master_host='192.168.1.251',
-> master_user='rep',
-> master_password='localhost',
-> master_log_file='mysql-bin.000001',
-> master_log_pos=0;
Query OK, 0 rows affected (0.02 sec)
mysql> show slave statusG
2. mysql2 192.168.1.251:
[root@mysql2 ~]# vim /etc/my.cnf
log-bin=mysql-bin
server-id = 251
[root@mysql2 ~]# service mysqld start
[root@mysql2 ~]# mysql
mysql> reset master;
mysql> grant replication slave, replication client on *.*
-> to rep@'192.168.1.%' identified by 'localhost';
mysql> flush privileges;
mysql> change master to
-> master_host='192.168.1.4',
-> master_user='rep',
-> master_password='localhost',
-> master_log_file='mysql-bin.000001',
-> master_log_pos=0;
Query OK, 0 rows affected (0.02 sec)
mysql> show slave statusG
3. mysql1,mysql2
mysql> slave start;
mysql> show slave statusG
4. 测试
5. 建立用于客户连接用户
mysql> grant ALL on *.* to admin@'192.168.1.%' identified by 'localhost';
mysql> flush privileges;
===========================================================================
生产环境其他常用设置
1、配置忽略权限库同步参数
binlog-ignore-db='information_schema mysql test'
2、从库备份开启binlog
log-slave-updates
log_bin = mysql-bin
expire_logs_days = 7
应用场景：级联复制或从库做数据备份。
3、从库只读read-only来实现
innodb_read_only = ON或1，或者innodb_read_only
结论：当用户权限中没有SUPER权限(ALL权限是包括SUPER的)时，从库的read-only生效！