基于mysqldump搭建gtid主从

1.如果master所有的binlog还在，安装slave后，直接change master 到master
原理是直接获取master所有的gtid并执行
优点是简单
缺点是如果binlog太多，数据完全同步需要的时间较长，并且需要master一开始就启用了GTID
总结：适用于master也是新建不久的情况

2.通过master或者其它slave的mysqldump备份搭建新的slave.
原理：备份时获取master的数据和这些数据对应的GTID，在Slave端跳过备份包含的GTID
优点是可以避免第一种方法中的不足
缺点操作相对复杂
总结：适用于拥有较大数据集的情况

3、percona xtrabackup
基于xtrabackup备份文件xtrabackup_binlog_info包含了GTID信息
做从库恢复后，需要手工设置：
set@@GLOBAL.GTID_PURGED='c8d960f1-83ca-11e5-a8eb-000c29ea831c:1-745497';
恢复后，执行change master to
缺点操作相对复杂
总结：适用于拥有较大数据集的情况

1、演示环境
mysql> system cat /etc/redhat-release
CentOS release 6.7 (Final)
mysql> show variables like 'version';
+---------------+------------+
| Variable_name | Value      |
+---------------+------------+
| version       | 5.7.12-log |
+---------------+------------+

主服务器：192.168.1.245:3306  server_id : 245
从服务器：192.168.1.247:3306  server_id : 247

--在主库端创建复制用户
mysql> GRANT REPLICATION SLAVE, REPLICATION CLIENT ON *.* TO 'repl'@'%' IDENTIFIED BY '123456'; 

2、直接使用change master(针对本文第一部分，第1小点情形)

此处省略基于gtid配置的参数描述，具体可以参考：配置MySQL GTID 主从复制
在从服务器端直接change master，如下：

SLAVE> show variables like 'server_id';
+---------------+-------+
| Variable_name | Value |
+---------------+-------+
| server_id     | 247   |
+---------------+-------+

Slave> CHANGE MASTER TO  
    -> MASTER_HOST='192.168.1.245',    
    -> MASTER_USER='repl',    
    -> MASTER_PASSWORD='123456',    
    -> MASTER_PORT=3306,    
    -> MASTER_AUTO_POSITION = 1;
Query OK, 0 rows affected, 2 warnings (0.12 sec)

Slave> start slave;
Query OK, 0 rows affected (0.01 sec)

Slave> start slave;
Query OK, 0 rows affected (0.01 sec)

Slave> show slave status \G
*************************** 1. row ***************************
               Slave_IO_State: Waiting for master to send event
                  Master_Host: 192.168.1.245
                  Master_User: repl
                  Master_Port: 3306
                Connect_Retry: 60
              Master_Log_File: node3-binlog.000001
          Read_Master_Log_Pos: 457
               Relay_Log_File: node5-relay-bin.000002
                Relay_Log_Pos: 676
        Relay_Master_Log_File: node3-binlog.000001
             Slave_IO_Running: Yes
            Slave_SQL_Running: Yes
              ...............

--主服务器端操作如下
Master> create database tempdb;
Query OK, 1 row affected (0.02 sec)

Master> use tempdb
Database changed
Master> create table t1(id int,ename varchar(20));
Query OK, 0 rows affected (0.09 sec)

Master> insert into t1 values(1,'leshami');
Query OK, 1 row affected (0.08 sec)

--从服务器端验证
Slave> select * from tempdb.t1;
+------+---------+
| id   | ename   |
+------+---------+
|    1 | leshami |
+------+---------+
1 row in set (0.01 sec)

3、基于mysqldump搭建gtid从库 
--准备环境，从库端执行
Slave> stop slave;          --停止重库
Query OK, 0 rows affected (0.01 sec)

Slave> reset slave all;     --重置主从配置信息
Query OK, 0 rows affected (0.02 sec)   

--准备环境，主库端执行  
Master> source sakila-db/sakila-schema.sql  --导入mysql自带的sakila数据库
Master> source sakila-db/sakila-data.sql    --填充数据   

--使用mysqldump导出数据库  
# mysqldump --all-databases --single-transaction --triggers --routines --events \
> --host=localhost --port=3306 --user=root --password=MyP@ssw0rd >/tmp/alldb.sql        

--导出的文件中已经包含了GTID_PURGED的信息
# grep GTID_PURGED /tmp/alldb.sql   
SET @@GLOBAL.GTID_PURGED='78336cdc-8cfb-11e6-ba9f-000c29328504:1-38';

--将备份文件copy到从服务器
# scp /tmp/alldb.sql 192.168.1.247:/tmp

-- 执行reset master，重置从服务器上的binlog
Slave> reset master;
Query OK, 0 rows affected (0.03 sec)

Slave> source /tmp/alldb.sql

Slave> show databases;    --此时tempdb已产生
+--------------------+
| Database           |
+--------------------+
| information_schema |
| mysql              |
| performance_schema |
| sakila             |
| sys                |
| tempdb             |
+--------------------+

--执行change master
Slave> CHANGE MASTER TO  
    -> MASTER_HOST='192.168.1.245',    
    -> MASTER_USER='repl',    
    -> MASTER_PASSWORD='123456',    
    -> MASTER_PORT=3306,    
    -> MASTER_AUTO_POSITION = 1;
Query OK, 0 rows affected, 2 warnings (0.06 sec)

Slave> start slave;
Query OK, 0 rows affected (0.00 sec)

Slave> show slave status \G
*************************** 1. row ***************************
               Slave_IO_State: Waiting for master to send event
                  Master_Host: 192.168.1.245
                  Master_User: repl
                  Master_Port: 3306
                Connect_Retry: 60
              Master_Log_File: node3-binlog.000001
          Read_Master_Log_Pos: 25637
               Relay_Log_File: node5-relay-bin.000002
                Relay_Log_Pos: 423
        Relay_Master_Log_File: node3-binlog.000001
             Slave_IO_Running: Yes
            Slave_SQL_Running: Yes

--主库端执行一些事务
Master> alter table tempdb.t1 modify ename varchar(50);
Query OK, 0 rows affected (0.05 sec)
Records: 0  Duplicates: 0  Warnings: 0

Master> insert into tempdb.t1 values(2,'http://blog.csdn.net/leshami');
Query OK, 1 row affected (0.02 sec)

--从库端验证结果
Slave> desc tempdb.t1;
+-------+-------------+------+-----+---------+-------+
| Field | Type        | Null | Key | Default | Extra |
+-------+-------------+------+-----+---------+-------+
| id    | int(11)     | YES  |     | NULL    |       |
| ename | varchar(50) | YES  |     | NULL    |       |
+-------+-------------+------+-----+---------+-------+
2 rows in set (0.00 sec)

Slave> select * from tempdb.t1;
+------+------------------------------+
| id   | ename                        |
+------+------------------------------+
|    1 | leshami                      |
|    2 | http://blog.csdn.net/leshami |
+------+------------------------------+