The recent reboot requirements at Amazon AWS have brought up the topic of automated restarts of MySQL database servers. I’d like to offer a safe solution for scripting the procedure. Let’s pretend that we’ve already confirmed there’s no production traffic. Here are the basic commands:

Stopping MySQL Server

Stop replication

mysqladmin -u root -p stop-slave

Make sure all in-memory stuff is saved to disk.

mysqladmin -u root -p flush-tables

Bring down the mysqld server

mysqladmin -u root -p shutdown

Starting MySQL Server

If mysqld isn’t started automatically when your host is booted, your distribution may have it’s own start script. However, you could always do the generic start:

nohup mysqld_safe &

Next, restart replication

mysqladmin -u root -p stop-slave