I had a problem that after 6+ months of replication running smoothly and nothing changing in terms of config, or the type of transactions being replicated the slaves (3) stopped with the error:
Got fatal error 1236: 'log event entry exceeded max_allowed_packet; Increase max_allowed_packet on master' from master when reading data from binary log
I increased the max_allowed_packet and then the slaves appeared to start and run, ie no error when I do SHOW SLAVE STATUS.
Slave_IO_State = Waiting for master to send event
Master_log_file = localhost-bin.188
Read_Master_Log_Pos = 90656760
Slave_IO_Running = Yes
Slave_SQL_Running = Yes
Last_Error = ""
Seconds_Behind_Master = 0
File = localhost-bin.191
Position = 79
So question is why do slaves say 0 seconds behind, but they are not at the same point as the master?