knowledgebase/linux/miscapplications/setting up mysql replication In the master server [mysqld] section of /etc/my.cnf you should have something like: {code:none} [mysqld] server-id=1 log-bin # so user changes do not replicate to the slaves replicate-ignore-db=mysql log-warnings log-slave-updates skip-bdb {code} On the master server run: {code:none} GRANT REPLICATION SLAVE ON *.* TO mysql_slave@slaveserverip identified by 'somepassword' {code} In the slave MySQL server's /etc/my.cnf file you should have something like: {code:none} server-id = 2 log-bin {code} Note that the server ids need to be unique for all the servers in your mysql replication cluster. On the master server run: {code}SHOW MASTER STATUS{code}. On the slave server's MySQL client use some of the values from that command to run the following: {code:none} CHANGE MASTER TO MASTER_HOST='masterserverip' , MASTER_USER='mysql_slave', MASTER_PASSWORD='somepassword' , MASTER_LOG_FILE='logfilenameperabove-bin.000003' , MASTER_LOG_POS=perabove; SLAVE START; {code} 2005-09-02 22:42:29.0 2005-09-02 22:41:20.0