Monitoring MySQL’s replication has always been a bit hit and miss, especially when trying to detect whether a slave is becoming overloaded or not. There’s been a few ways to do this in the past:
- Monitor the Seconds_behind_master variable from SHOW SLAVE STATUS, which is the difference between the timestamp passed down from the master that the event was executed, to the time that the SQL thread started executing the event.
- Use something like mk-heartbeat, which operates in the same way as Seconds_Behind_Master (trying to show you the actual time difference between the master and slave),