MySQL示例DTID主从原理解析
导读
目录
1.GTID基本概念2.GTID优点3.GTID的工作原理4.GTID比传统复制的优势5.启动的方法6.GTID(一主一从)配置6.1环境：6.2在主库上给从库授权：6.3确保数据一致操作6.4配置主库6.5配置从库6.6配置主从复制7.GTID（一主俩从）8.GTID（俩主一从）1.最新环境2.所有服务器均关闭防火墙或者放行防火墙3.授权连接master01库授权普通用户slave进行连接master02授权普通用户slave进行连接4.分别进行配置文件修改5.分别重启6.在进行GTID多主一从配置前，先引入一个概念7.slave从库以root用户登录进行GTID配置GTID(俩主一从)测试：slave相关命令：解决问题
1.GTID基本概念
MySQL 5.6.5开始支持的，全局事务标识符（GTID（Global Transaction ID））是创建的唯一标识符，并与在源（主）服务器上提交的每个事务相关联。
此标识符不但是唯一的，而且在给定复制设置中的所有服务器上都是唯一的。
所有交易和所有GTID之间都有一对一的映射关系 。
它由服务器ID以及事务ID组合而成。
这个全局事务ID不仅仅在原始服务器上唯一，在所有存在主从关系 的mysql服务器上也是唯一的。
正是因为这样一个特性使得mysql的主从复制变得更加简单，以及数据库一致性更可靠。
一个GTID在一个服务器上只执行一次，避免重复执行导致数据混乱或者主从不一致。
2.GTID优点
保证同一个事务在某slave上绝对只执行一次，没有执行过的gtid事务总是会被执行。
不用像传统复制那样保证binlog的坐标准确，因为根本不需要binlog以及坐标。
故障转移到新的master的时候很方便，简化了很多任务。
很容易判断master和slave的数据是否一致。只要master上提交的事务在slave上也提交了，那么一定是一致的。
3.GTID的工作原理
1.当一个事务在主库端执行并提交时，产生GTID，一同记录到binlog日志中。
2.binlog传输到slave,并存储到slave的relaylog后，读取这个GTID的这个值设置gtid_next变量，即告诉Slave，下一个要执行的GTID值。
3、sql线程从relay log中获取GTID，然后对比slave端的binlog是否有该GTID。
4、如果有记录，说明该GTID的事务已经执行，slave会忽略。
5、如果没有记录，slave就会执行该GTID事务，并记录该GTID到自身的binlog，在读取执行事务前会先检查其他session持有该GTID，确保不被重复执行。
6、在解析过程中会判断是否有主键，如果有就用二级索引，如果没有就用全部扫描。
4.GTID比传统复制的优势
1.更简单的实现故障转移（failover），不需要找log_file,log_pos
2.更简单的搭建主从复制
3.更加安全
4.GTID是连续没有空洞的，因此主数据库发生冲突时，可以添加空事件的方式进行跳过
5.启动的方法
方法一：如果是新搭建的服务器，直接启动即可 方法二：如果是以及跑的服务器，需要重启一下mysql server
启动前，先关闭master的写入，保证master端和slave端数据保持同步，所有slave需要加上skip_slave_start=1的配置参数，避免启动后还是使用之前的复制协议
6.GTID(一主一从)配置
6.1环境：
centos8.0 ip：192.168.136.239 有数据 hostname：mysql01
centos8.0 ip：192.168.136.219 无数据 hostname：mysql02
#二进制安装以及mysql自启动服务略
6.2在主库上给从库授权：
mysql> grant replication slave on *.* to 'slave'@'192.168.136.219' identified by 'slave'; Query OK, 0 rows affected, 1 warning (0.00 sec) mysql> flush privileges; Query OK, 0 rows affected (0.00 sec) #俩服务器均关闭防火墙 [root@mysql01 ~]# systemctl stop firewalld [root@mysql01 ~]# setenforce 0 [root@mysql02 ~]# systemctl stop firewalld [root@mysql02 ~]# setenforce 0 从库测试连接： [root@mysql02 ~]# mysql -u slave -p'slave' -h192.168.136.239 Type 'help;' or '\h' for help. Type '\c' to clear the current input statement. mysql>
6.3确保数据一致操作
1.对主库进行锁表 mysql> flush tables with read lock; 2.对主库进行全备 [root@mysql01 ~]# mysqldump -uroot -A > /clq/all-databases-20210519.sql 3.拷贝到从库主机上去 [root@mysql01 ~]# scp /clq/all-databases-20210519.sql root@192.168.136.219:/backup/ [root@mysql02 backup]# ll -rw-r--r--. 1 root root 873527 5月 19 16:40 all-databases-20210519.sql 4.从库上进行主库的恢复 [root@mysql02 backup]# mysql -uroot -pHuawei0917@ < all-databases-20210519.sql
6.4配置主库
[mysqld] basedir = /usr/local/mysql datadir = /opt/data socket = /tmp/mysql.sock port = 3306 user = mysql pid-file = /opt/data/mysql.pid skip-name-resolve #skip-grant-tables log-bin = master_bin #开启主库日志 server-id = 10 #服务唯一标识id gtid-mode = on #GTID模式开启 enforce_gtid_consistency = on #强制gtid模式一致性 log-slave-updates = 1 #从库允许更新日志，同步操作日志 binlog_format = row #binlog日志格式为行格式, 默认是mixed混合模式 skip_slave_start = 1 #跳过从库开启，以主库开始开启 #重启 systemctl restart mysqld
6.5配置从库
[root@mysql02 data]# cat /etc/my.cnf [mysqld] basedir = /usr/local/mysql datadir = /opt/data socket = /tmp/mysql.sock port = 3306 user = mysql pid-file = /opt/data/mysql.pid skip-name-resolve #skip-grant-tables gtid_mode=on enforce_gtid_consistency=on server-id=20 log-bin=slave_binlog #开启从库日志 log_slave-updates=1 #从库允许更新 binlog_format=row #格式为行 skip-slave_start=1 #重启 systemctl restart mysqld
查看gtid状态情况
mysql> show variables like '%gtid%'; +----------------------------------+-----------+ | Variable_name | Value | +----------------------------------+-----------+ | binlog_gtid_simple_recovery | ON | | enforce_gtid_consistency | ON | | gtid_executed_compression_period | 1000 | | gtid_mode | ON | | gtid_next | AUTOMATIC | | gtid_owned | | | gtid_purged | | | session_track_gtids | OFF | +----------------------------------+-----------+ 8 rows in set (0.00 sec)
6.6配置主从复制
#从库上root登录配置 #help change master to 可以查看帮助文档实例 mysql> change master to -> master_host='192.168.136.239', -> master_user='slave', -> master_password='slave', -> master_port=3306, #主库端口 -> master_auto_position=1; #位置 #master_use_gtid = current_pos Query OK, 0 rows affected, 2 warnings (0.01 sec) mysql> start slave; Query OK, 0 rows affected (0.00 sec) mysql> show slave status\G; Slave_IO_Running: Connecting Slave_IO_Running: Yes Slave_SQL_Running: Yes 保证系统一致性 授权一致性
（一主一从GTID）测试：
主库创建一个数据库test，进行测试查看
从库创建一个数据库test02，进行测试查看
#主库创建一个test数据库 mysql> create database test; mysql> show databases; +--------------------+ | Database | +--------------------+ | information_schema | | mysql | | performance_schema | | sys | | test | +--------------------+ #从库上查看同步情况 mysql> show databases; +--------------------+ | Database | +--------------------+ | information_schema | | mysql | | performance_schema | | sys | | test | +--------------------+ 6 rows in set (0.00 sec) #从库创建test02库 mysql> create database test02; Query OK, 1 row affected (0.00 sec) #主库上查看 mysql> show databases; +--------------------+ | Database | +--------------------+ | information_schema | | mysql | | performance_schema | | sys | #是没有test02库的 | test | +--------------------+ 5 rows in set (0.00 sec)
小结：主库上的数据操作会同步到从库上面去，而从库上的数据操作与主库没联系
7.GTID（一主俩从）
第三台mysql连接的话，相应配置
第3台mysql ，版本:centos8 ip:192.168.136.230 主机名:mysql03
[root@mysql03 ~]# cat /etc/my.cnf [mysqld] basedir = /usr/local/mysql datadir = /opt/data socket = /tmp/mysql.sock port = 3306 user = mysql pid-file = /opt/data/mysql.pid skip-name-resolve #skip-grant-tables # replication config log-bin = master_bin server-id = 21 #id必须与之前不同 gtid-mode = on enforce-gtid-consistency = on log-slave-updates = 1 binlog-format = row skip-slave-start = 1 #查看gtid情况 mysql> show variables like '%gtid%'; +----------------------------------+-----------+ | Variable_name | Value | +----------------------------------+-----------+ | binlog_gtid_simple_recovery | ON | | enforce_gtid_consistency | ON | | gtid_executed_compression_period | 1000 | | gtid_mode | ON | | gtid_next | AUTOMATIC | | gtid_owned | | | gtid_purged | | | session_track_gtids | OFF | +----------------------------------+-----------+ #由于之前只权限了一个ip，此刻在mysql01主数据库上再授权一个ip mysql> grant replication slave on *.* to 'slave'@'192.168.136.230' identified by 'slave'; Query OK, 0 rows affected, 1 warning (0.00 sec) mysql> flush privileges; Query OK, 0 rows affected (0.00 sec) #测试连接 [root@mysql ~]# mysql -uslave -pslave -h192.168.136.239 mysql: [Warning] Using a password on the command line interface can be insecure. Welcome to the MySQL monitor. Commands end with ; or \g. Your MySQL connection id is 17 Server version: 5.7.33-log MySQL Community Server (GPL) Copyright (c) 2000, 2021, Oracle and/or its affiliates. Oracle is a registered trademark of Oracle Corporation and/or its affiliates. Other names may be trademarks of their respective owners. Type 'help;' or '\h' for help. Type '\c' to clear the current input statement. mysql> #mysql03从库上root用户连接进行相应配置 [root@mysql03 ~]# mysql -uroot -p1 mysql> change master to -> master_host='192.168.136.239', #主库ip -> master_user='slave', #主库授权的普通用户 -> master_password='slave', -> master_port=3306, #主库端口 -> master_auto_position=1; #位置从1开始同步 #也可以查看帮助进行配置 mysql> help change master to; CHANGE MASTER TO MASTER_HOST='source2.example.com', MASTER_USER='replication', MASTER_PASSWORD='password', MASTER_PORT=3306, MASTER_LOG_FILE='source2-bin.001', MASTER_LOG_POS=4, MASTER_CONNECT_RETRY=10; URL: https://dev.mysql.com/doc/refman/5.7/en/change-master-to.html #开启 mysql> start slave; mysql> show slave status\G; *************************** 1. row *************************** Slave_IO_State: Waiting for master to send event Master_Host: 192.168.136.239 Master_User: slave Master_Port: 3306 Connect_Retry: 60 Master_Log_File: master_bin.000002 Read_Master_Log_Pos: 2172 Relay_Log_File: mysql-relay-bin.000002 Relay_Log_Pos: 2387 Relay_Master_Log_File: master_bin.000002 Slave_IO_Running: Yes Slave_SQL_Running: Yes #显示俩个yes则运行成功！ #mysql03查看数据库，数据库内容也同步成功 mysql> show databases; +--------------------+ | Database | +--------------------+ | information_schema | | mysql | | performance_schema | | sys | | test | +--------------------+ 5 rows in set (0.00 sec)
8.GTID（俩主一从）
1.最新环境
| 版本 | ip | 主机名 | 身份 |
|---|---|---|---|
| centos8 | 192.168.136.239 | master01 | 主库 |
| centos8 | 192.168.136.219 | master02 | 主库 |
| centos8 | 192.168.136.230 | slave | 从库 |
2.所有服务器均关闭防火墙或者放行防火墙
[root@master01 ~]# systemctl stop firewalld [root@master01 ~]# systemctl disable firewalld [root@master02 ~]# systemctl stop firewalld [root@master02 ~]# systemctl disable firewalld [root@slave ~]# systemctl stop firewalld [root@slave ~]# systemctl disable firewalld
3.授权连接
master01库授权普通用户
mysql> grant replication slave on *.* to 'user'@'192.168.136.%' identified by 'user';
slave进行连接
[root@slave ~]# mysql -uuser -p'user' -h192.168.136.239 mysql: [Warning] Using a password on the command line interface can be insecure. Welcome to the MySQL monitor. Commands end with ; or \g. Your MySQL connection id is 5 Server version: 5.7.33 MySQL Community Server (GPL) Copyright (c) 2000, 2021, Oracle and/or its affiliates. Oracle is a registered trademark of Oracle Corporation and/or its affiliates. Other names may be trademarks of their respective owners. Type 'help;' or '\h' for help. Type '\c' to clear the current input statement.
master02授权普通用户
mysql> grant replication slave on *.* to 'app'@'192.168.136.%' identified by 'app'; Query OK, 0 rows affected, 1 warning (0.01 sec) mysql> flush privileges; Query OK, 0 rows affected (0.00 sec)
slave进行连接
[root@slave ~]# mysql -uapp -papp -h192.168.136.219 mysql: [Warning] Using a password on the command line interface can be insecure. Welcome to the MySQL monitor. Commands end with ; or \g. Your MySQL connection id is 3 Server version: 5.7.33 MySQL Community Server (GPL) Copyright (c) 2000, 2021, Oracle and/or its affiliates. Oracle is a registered trademark of Oracle Corporation and/or its affiliates. Other names may be trademarks of their respective owners. Type 'help;' or '\h' for help. Type '\c' to clear the current input statement. mysql>
4.分别进行配置文件修改
#master01主机： [root@master01 ~]# cat /etc/my.cnf [mysqld] basedir = /usr/local/mysql datadir = /opt/data socket = /tmp/mysql.sock port = 3306 user = mysql pid-file = /opt/data/mysql.pid skip-name-resolve skip-grant-tables log-bin = master_bin server-id = 10 gtid-mode = on enforce-gtid-consistency = on log-slave-updates = 1 binlog-format = row skip-slave-start = 1 #master02主机 [mysqld] basedir = /usr/local/mysql datadir = /opt/data socket = /tmp/mysql.sock port = 3306 user = mysql pid-file = /opt/data/mysql.pid skip-name-resolve #replication config log-bin = master_bin server-id = 11 gtid-mode = on enforce-gtid-consistency = on log-slave-updates = 1 binlog-format = row skip-slave-start = 1 #slave主机 [mysqld] basedir = /usr/local/mysql datadir = /opt/data socket = /tmp/mysql.sock port = 3306 user = mysql pid-file = /opt/data/mysql.pid skip-name-resolve log-bin = slave_bin server-id = 13 gtid-mode = on enforce-gtid-consistency = on log-slave-updates = 1 binlog-format = row skip-slave-start = 1
5.分别重启
[root@master01 ~]# systemctl restart mysqld [root@master02 ~]# systemctl restart mysqld [root@slave ~]# systemctl restart mysqld
6.在进行GTID多主一从配置前，先引入一个概念
channel(频道)：每一个channel都是一个独立的slave服务，都有一个IO_THREAD和SQL_THREAD，原理和普通复制一样，只是需要在change master to语句后面使用FOR Channel来进行区分slave
在使用channel时需要将从库的master-info-repository、relay-log-info-repository设置为table，否则会报错。
将信息存储库设置为table格式
方式一（mysql内设置）:
set global master_info_repository='table';
set global relay_log_info_repository='table';
方式二（/etc/my.cnf内设置）:
3.在my.cnf中设置
master_info_repository = TABLE
relay_log_info_repository = TABLE
#检查是否更改成功
mysql> show variables where variable_name in ('relay_log_info_repository','master_info_repository');
+---------------------------+-------+
| Variable_name | Value |
+---------------------------+-------+
| master_info_repository | TABLE |
| relay_log_info_repository | TABLE |
+---------------------------+-------+7.slave从库以root用户登录进行GTID配置
#slave从库上配置俩个主库GTID复制 mysql> change master to -> master_host='192.168.136.219', #mysql02主库ip -> master_user='app', #mysql02主库授权的普通用户 -> master_password='app', #mysql02主库授权的普通用户密码 -> master_port=3306, #主库端口 -> master_auto_position=1 for channel 'master01'; #位置从1开始同步,并且第一个slave取名master01 mysql> change master to -> master_host='192.168.136.239', #mysql01主库ip -> master_user='user', -> master_password='user', -> master_port=3306, #主库端口 -> master_auto_position=1 for channel 'master02'; #位置从1开始同步,并且第一个slave取名master01 #查看俩个slave状态 mysql> show slave status\G; *************************** 1. row *************************** Slave_IO_State: Master_Host: 192.168.136.219 Master_User: app Master_Port: 3306 Connect_Retry: 60 Master_Log_File: Read_Master_Log_Pos: 4 Relay_Log_File: slave02-relay-bin-master1.000001 Relay_Log_Pos: 4 Relay_Master_Log_File: Slave_IO_Running: No Slave_SQL_Running: No #都是关闭的 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: 0 Relay_Log_Space: 154 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: NULL 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: 0 Master_UUID: Master_Info_File: mysql.slave_master_info SQL_Delay: 0 SQL_Remaining_Delay: NULL Slave_SQL_Running_State: Master_Retry_Count: 86400 Master_Bind: Last_IO_Error_Timestamp: Last_SQL_Error_Timestamp: Master_SSL_Crl: Master_SSL_Crlpath: Retrieved_Gtid_Set: Executed_Gtid_Set: b4326a77-0a31-11ec-a991-000c298d3571:1-2, d68b404d-0a35-11ec-9df1-000c29581959:1 Auto_Position: 1 Replicate_Rewrite_DB: Channel_Name: master1 Master_TLS_Version: *************************** 2. row *************************** Slave_IO_State: Master_Host: 192.168.136.239 Master_User: user Master_Port: 3306 Connect_Retry: 60 Master_Log_File: Read_Master_Log_Pos: 4 Relay_Log_File: slave02-relay-bin-master2.000001 Relay_Log_Pos: 4 Relay_Master_Log_File: Slave_IO_Running: No Slave_SQL_Running: No 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: 0 Relay_Log_Space: 154 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: NULL 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: 0 Master_UUID: Master_Info_File: mysql.slave_master_info SQL_Delay: 0 SQL_Remaining_Delay: NULL Slave_SQL_Running_State: Master_Retry_Count: 86400 Master_Bind: Last_IO_Error_Timestamp: Last_SQL_Error_Timestamp: Master_SSL_Crl: Master_SSL_Crlpath: Retrieved_Gtid_Set: Executed_Gtid_Set: b4326a77-0a31-11ec-a991-000c298d3571:1-2, d68b404d-0a35-11ec-9df1-000c29581959:1 Auto_Position: 1 Replicate_Rewrite_DB: Channel_Name: master2 Master_TLS_Version: 2 rows in set (0.00 sec) #开启俩个slave mysql> start slave; #再次查看状态
GTID(俩主一从)测试：
#master01主库创建一个test数据库 mysql> create database test; Query OK, 1 row affected (0.00 sec) mysql> show databases; +--------------------+ | Database | +--------------------+ | information_schema | | mysql | | performance_schema | | sys | | test | +--------------------+ 5 rows in set (0.00 sec) #master02主库上查看 mysql> show databases; +--------------------+ | Database | +--------------------+ | information_schema | | mysql | | performance_schema | | sys | #没有内容 +--------------------+ 4 rows in set (0.00 sec) #slave从库查看 mysql> show databases; +--------------------+ | Database | +--------------------+ | information_schema | | mysql | | performance_schema | | sys | | test | #已经同步了test库 +--------------------+ 5 rows in set (0.00 sec) #mysql02主库创建一个RHCA数据库 mysql> create database RHCA; Query OK, 1 row affected (0.01 sec) mysql> show databases; +--------------------+ | Database | +--------------------+ | information_schema | | RHCA | | mysql | | performance_schema | | sys | +--------------------+ 5 rows in set (0.00 sec) #slave从库 mysql> show databases; +--------------------+ | Database | +--------------------+ | information_schema | | RHCA | | mysql | | performance_schema | | sys | #有了mysql01主库的test库和mysql02的RHCA的库 | test | +--------------------+ 6 rows in set (0.00 sec)
slave相关命令：
show slave status； //查看全部slave状态
show slave status for channel ‘naem'; //查看单个slave状态
reset slave; #重置全部slave
reset slave for channel ‘master1'; #重置单个slave
stop slave for channel ‘master1'; #暂停单个slave
start slave for channel ‘master1'; #开启单个slave
虽然我在做的过程没有遇到错误，但是下面这个是最最容易出现的错误
配置完开启slave出现报错
mysql> start slave; ERROR 1872 (HY000): Slave failed to initialize relay log info structure from the repository
解决问题
由于mysql.slave_relay_log_info表中保留了以前的复制信息，导致新从库启动时无法找到对应文件，那么我们清理掉该表中的记录即可
mysql> reset slave; Query OK, 0 rows affected (0.00 sec)
猜你喜欢
评论(0)
