Master-slave replication in MySQL

Master slave replication refers to the transfer of the master database DDL and DML The operation is passed to the database server through binary log , Then re execute the logs on the slave database （ It's also called redoing ）, In this way, the data of the slave database and the master database can be synchronized .

MySQL One master database can be copied to multiple slave databases at the same time , The slave library can also be used as the master library of other slave servers , Achieve Chain replication .

MySQL The copied points mainly include the following three aspects ：
1、 There's something wrong with the main database , You can quickly switch to providing services from the library .
2、 Read and write separation , Reduce the access pressure of the main library .
3、 You can perform a backup from the library , To avoid affecting the main library service during backup .

Master slave replication principle ：
1、 When the main database commits the transaction , The data changes will be recorded in the binary log file Binlog in .
2、 Read the binary log file of the main library from the library Binlog, Write to the continuation log of the slave Library Relay Log.
3、 Resume the events in the log from the library redo , Will change the data that reflects itself .

build

• Get ready
  When you have two servers ready , Install... In the above two servers respectively MySQL, And complete the basic initialization preparation .

# Develop specified 3306 Port number 
firewall-cmd --zone=public --add-port=3306/tcp -permanent
firewall-cmd -reload

• Main library configuration
  1、 Modify the configuration file /etc/my.cnf

#mysql service ID, Ensure that the entire cluster environment is unique , Value range ：1-2^32-1, The default is 1
server-id=1
# Is it read-only ,1 Read only ,0 For reading and writing 
read-only=0
# Ignored data , A database that does not need to be synchronized 
#binlog-ignore-db=mysql
# Specify a synchronized database 
#binlog-do-db=db_name

2、 restart MySQL The server
systemctl restart mysqld
3、 Sign in mysql, Create a remote connection account , And grant master-slave replication permission

# establish itcast user , And set the password , The user can connect to the server from any host MySQL service 
create user 'itcast'@'%' identified with mysql_native_password by 'password';
# by 'itcast'@'%' The user assigns master-slave copy permission 
grant replication slave on *.* to 'itcast'@'%';

4、 Passing instructions , View binary log coordinates
show master status;
Field meaning description ：
file： From which log file to start pushing log files
position： Where to start pushing logs
binlog_ignore_db： Specify a database that does not need to be synchronized

• Configuration from library
  1、 Modify the configuration file /et/my.cnf

#mysql service ID, Ensure that the entire cluster environment is unique , Value range ：1~2^32-1, It is different from the main library 
server-id=2
# Is it read-only ,1 Read only ,0 For reading and writing 
read-only=1

2、 Restart MySQL service
systemctl restart mysqld

3、 Sign in mysql, Set master library configuration
change replication source to source_host='xxx.xxx',source_user='xxx',source_password='xxx',source_log_file='xxx',source_log_pos='xxx';
The above is 8.0.23 Chinese grammar . If mysql yes 8.0.23 Previous version , The implementation is as follows SQL：
change master to master_host='xxx.xxx.xxx.xxx',master_user='xxx',master_password='xxx',master_log_file='xxx',master_log_pos=xxx;

4、 Turn on synchronous operation

start replica; #8.0.22 after 
start slave;   #8.0.22 Before 

5、 Check the master-slave synchronization status

show replica status;	#8.0.22 after 
show slave status;      #8.0.22 Before