Bearing in mind the number on the image, let’s see how the process works:
- Item 1 in the image represents the clients executing queries on master, note that the master is capable to handle multiple simultaneous connections (it can be configurable by
- Master process these queries, and save it to his binary log(item number 2 in the image), then it can later on be replied on the slave.
- The slave has 2 threads reubuntu mount ext4 partitionsponsible to deal with replication :
IO_THREAD– is responsible to connect to the Master, ask for new transaction (item number 3 in the image) and save them to his own log (relay log, item number 4 in the image).
SQL_THREAD– is responsible for read all new transaction from relay log(item number 5 in the image), and apply them to the database server(item number 6 in the image) .