I’ve started playing around with MHA and I should say it works great for manual or even automated failover of master-slave replication servers.
Steps 1 and 2 are solid.
Step 3. you’d want to do, STOP SLAVE; RESET SLAVE ALL;
as it is 5.6
Step 4. set global read_only=0;
Step 5. these are optional and read_only can be changed online.
Step 6. SHOW MASTER STATUS; on old slave/new master
Step 7. CHANGE MASTER TO … (use log file and position from #6)
Step 8. set global read_only=1; on new slave/old master
Step 8. start slave; show slave status\G on new slave/old master.
Again with MHA ([url]https://code.google.com/p/mysql-master-ha/wiki/Overview[/url]), it would be as easy as :
shell> masterha_master_switch --master_state=alive --conf=/etc/app1.cnf --new_master_host=old_slave --orig_master_is_new_slave