That's why! Why can we continue the master-slave replication relationship after performing stop slave and then start slave?

Write it on the front:

We all know that Master is full of a binlog, or Master manually executes flush logs. SLave can continue to receive data from Master for synchronization without doing anything. I don't know if my friends have considered it. Why is this?

The reason is:

In fact, executing stop slave means closing the I/O thread (stop slave IO_THREAD) and SQL thread (stop slave SQL_THREAD) respectively. The I/O thread will maintain the update of master.info information, and the SQL thread will maintain the update of relay-log.info information. When executing start slave, I/O thread and SQL thread will continue to execute according to master.info and relay-log.info information, so replication can continue.

Here's a demonstration:

Execute on Master host:

mysql> show master status\G;

*************************** 1. row ***************************

File: mysql-bin.000008

Position: 120

Executed on Slave host

cat /home/data/mysql3306/master.info

mysql-bin.000008

120

192.168.32.3

Now manually scroll through the binlog log once in Master

mysql> flush logs;

mysql> show master status\G;

*************************** 1. row ***************************

File: mysql-bin.000009

Position: 120

Executed on Slave host

cat /home/data/mysql3306/master.info

23

mysql-bin.000009

120

192.168.32.3

As you can see, the Slave host copies Master's data exactly according to the master.info file.