Sometime ago we configured MYSQL replication on our main web server to another CentOS 6 server. All is working perfectly and we can see the replication is working fine.
My question is how to safely reboot either master or slave server without breaking the replication and thus having to reconfigure?
If a slave server goes down, replication will stop but it won't require you to reconfigure master/slave replication. The master will simply wait for the slave to become available again at which point pending transactions will be synchronised to the slave. If you need to reboot the master server you will have to promote the slave to master. This can be achieved as follows:
On the master: FLUSH LOGS;
On the slave: STOP SLAVE; RESET MASTER; CHANGE MASTER TO MASTER_HOST='';
