详解MySQL主从复制实战 - 基于GTID的复制

 基于gtid的复制

简介

基于gtid的复制是mysql 5.6后新增的复制方式.

gtid (global transaction identifier) 即全局事务id, 保证了在每个在主库上提交的事务在集群中有一个唯一的id.

在原来基于日志的复制中, 从库需要告知主库要从哪个偏移量进行增量同步, 如果指定错误会造成数据的遗漏, 从而造成数据的不一致.

而基于gtid的复制中, 从库会告知主库已经执行的事务的gtid的值, 然后主库会将所有未执行的事务的gtid的列表返回给从库. 并且可以保证同一个事务只在指定的从库执行一次.

实战

1、在主库上建立复制账户并授予权限

基于gtid的复制会自动地将没有在从库执行的事务重放, 所以不要在其他从库上建立相同的账号. 如果建立了相同的账户, 有可能造成复制链路的错误.

mysql> create user 'repl'@'172.%' identified by '123456';

注意在生产上的密码必须依照相关规范以达到一定的密码强度, 并且规定在从库上的特定网段上才能访问主库.

mysql> grant replication slave on *.* to 'repl'@'172.%';

查看用户

mysql> select user, host from mysql.user;
+-----------+-----------+
| user  | host  |
+-----------+-----------+
| prontera | %   |
| root  | %   |
| mysql.sys | localhost |
| root  | localhost |
+-----------+-----------+
4 rows in set (0.00 sec)

查看授权

mysql> show grants for repl@'172.%';
+--------------------------------------------------+
| grants for repl@172.%       |
+--------------------------------------------------+
| grant replication slave on *.* to 'repl'@'172.%' |
+--------------------------------------------------+
1 row in set (0.00 sec)

2、配置主库服务器

[mysqld]
log_bin = /var/log/mysql/mysql-bin
log_bin_index = /var/log/mysql/mysql-bin.index
binlog_format = row
server_id = 101
gtid_mode = on
enforce_gtid_consistency = on
#log_slave_updates = on

note: 把日志与数据分开是个好习惯, 最好能放到不同的数据分区

enforce_gtid_consistency 强制gtid一致性, 启用后以下命令无法再使用

create table ... select ...

mysql> create table dept select * from departments;
error 1786 (hy000): statement violates gtid consistency: create table ... select.

因为实际上是两个独立事件, 所以只能将其拆分先建立表, 然后再把数据插入到表中

create temporary table

事务内部不能创建临时表

mysql> begin;
query ok, 0 rows affected (0.00 sec)

mysql> create temporary table dept(id int);
error 1787 (hy000): statement violates gtid consistency: create temporary table and drop temporary table can only be executed outside transactional context. these statements are also not allowed in a function or trigger because functions and triggers are also considered to be multi-statement transactions.

同一事务中更新事务表与非事务表(myisam)

mysql> create table `dept_innodb` (id int(11) unsigned not null primary key auto_increment);
query ok, 0 rows affected (0.04 sec)

mysql> create table `dept_myisam` (id int(11) unsigned not null primary key auto_increment) engine = `myisam`;
query ok, 0 rows affected (0.03 sec)

mysql> begin;
query ok, 0 rows affected (0.00 sec)

mysql> insert into dept_innodb(id) value(1);
query ok, 1 row affected (0.00 sec)

mysql> insert into dept_myisam(id) value(1);
error 1785 (hy000): statement violates gtid consistency: updates to non-transactional tables can only be done in either autocommitted statements or single-statement transactions, and never in the same statement as updates to transactional tables.

所以建议选择innodb作为默认的数据库引擎.

log_slave_updates 该选项在mysql 5.6版本时基于gtid的复制是必须的, 但是其增大了从服务器的io负载, 而在mysql 5.7中该选项已经不是必须项

3、配置从库服务器

master_info_repository 与relay_log_info_repository

在mysql 5.6.2之前, slave记录的master信息以及slave应用binlog的信息存放在文件中, 即master.info与relay-log.info. 在5.6.2版本之后, 允许记录到table中. 对应的表分别为mysql.slave_master_info与mysql.slave_relay_log_info, 且这两个表均为innodb引擎表.

[mysqld]
log_bin = /var/log/mysql/mysql-bin
log_bin_index = /var/log/mysql/mysql-bin.index
server_id = 102
# slaves
relay_log  = /var/log/mysql/relay-bin
relay_log_index = /var/log/mysql/relay-bin.index
relay_log_info_file = /var/log/mysql/relay-bin.info
enforce_gtid_consistency = on
log_slave_updates = on
read_only = on
master_info_repository = table
relay_log_info_repository = table

4、从库数据初始化 - [optional]

先在主库上备份数据

mysqldump --single-transaction --master-data=2 --triggers --routines --all-databases --events -u root -p > backup.sql

—master-data=2 该选项将当前服务器的binlog的位置和文件名追加到输出文件中(show master status). 如果为1, 将偏移量拼接到change master 命令. 如果为2, 输出的偏移量信息将会被注释。

--all-databases 因为基于gtid的复制会记录全部的事务, 所以要构建一个完整的dump这个选项是推荐的

常见错误

当从库导入sql的时候出现

error 1840 (hy000) at line 24: @@global.gtid_purged can only be set when @@global.gtid_executed is empty.

此时进入从库的mysql command line, 使用reset master即可

5、启动基于gtid的复制

现有master@172.20.0.2和slave@172.20.0.3, 并且已经通过mysqldump将数据同步至从库slave中. 现在在从服务器slave上配置复制链路

mysql> change master to master_host='master', master_user='repl', master_password='123456', master_auto_position=1;
query ok, 0 rows affected, 2 warnings (0.06 sec)

启动复制

mysql> start slave;

启动成功后查看slave的状态

mysql> show slave status\g
*************************** 1. row ***************************
    slave_io_state: queueing master event to the relay log
     master_host: master
     master_user: repl
     master_port: 3306
    connect_retry: 60
    master_log_file: mysql-bin.000002
   read_master_log_pos: 12793692
    relay_log_file: relay-bin.000002
    relay_log_pos: 1027
  relay_master_log_file: mysql-bin.000002
    slave_io_running: yes
   slave_sql_running: yes
    replicate_do_db:
   replicate_ignore_db:
   replicate_do_table:
  replicate_ignore_table:
  replicate_wild_do_table:
 replicate_wild_ignore_table:
     last_errno: 0
     last_error:
     skip_counter: 0
   exec_master_log_pos: 814
    relay_log_space: 12794106
    until_condition: none
    until_log_file:
    until_log_pos: 0
   master_ssl_allowed: no
   master_ssl_ca_file:
   master_ssl_ca_path:
    master_ssl_cert:
   master_ssl_cipher:
    master_ssl_key:
  seconds_behind_master: 5096
master_ssl_verify_server_cert: no
    last_io_errno: 0
    last_io_error:
    last_sql_errno: 0
    last_sql_error:
 replicate_ignore_server_ids:
    master_server_id: 101
     master_uuid: a9fd4765-ec70-11e6-b543-0242ac140002
    master_info_file: mysql.slave_master_info
     sql_delay: 0
   sql_remaining_delay: null
  slave_sql_running_state: reading event from the relay log
   master_retry_count: 86400
     master_bind:
  last_io_error_timestamp:
  last_sql_error_timestamp:
    master_ssl_crl:
   master_ssl_crlpath:
   retrieved_gtid_set: a9fd4765-ec70-11e6-b543-0242ac140002:1-39
   executed_gtid_set: a9fd4765-ec70-11e6-b543-0242ac140002:1-4
    auto_position: 1
   replicate_rewrite_db:
     channel_name:
   master_tls_version:
1 row in set (0.00 sec)

当slave_io_running, slave_sql_running为yes,

且slave_sql_running_state 为slave has read all relay log; waiting for more updates时表示成功构建复制链路

6、总结

优点

1. 因为不用手工设置日志偏移量, 可以很方便地进行故障转移
2. 如果启用log_slave_updates那么从库不会丢失主库上的任何修改

缺点

1. 对执行的sql有一定限制
2. 仅支持mysql 5.6之后的版本, 而且不建议使用早期5.6版本

以上就是本文的全部内容，希望对大家的学习有所帮助，也希望大家多多支持。