I've scoured the internet but can't seem to find whether or not the mysql.slave_relay_log_info is replicated down a chain or around in a circular topology.
The enterprise monitor agent - from what I understand - disables writing to the bin log when it adds host identifier information to mysql.inventory.
It wouldn't make any sense for a master (that is a slave and has a slave) to disable this with every update to mysql.slave_relay_log_info.
So, is it ignored on apply? Anyone have the answer?
Also it isn't stated anywhere what happens when you go from relay_log_info_repository=FILE to relay_log_info_repository=TABLE.
Is it copying the data from the file into the table? When you go in reverse, does it copy the data from the table and put it into the file? Do you have to do this while connected - modifying a global variable - then also putting it in the my.cnf OR can you make the change in the my.cnf and just restart the server?
The enterprise monitor agent - from what I understand - disables writing to the bin log when it adds host identifier information to mysql.inventory.
It wouldn't make any sense for a master (that is a slave and has a slave) to disable this with every update to mysql.slave_relay_log_info.
So, is it ignored on apply? Anyone have the answer?
Also it isn't stated anywhere what happens when you go from relay_log_info_repository=FILE to relay_log_info_repository=TABLE.
Is it copying the data from the file into the table? When you go in reverse, does it copy the data from the table and put it into the file? Do you have to do this while connected - modifying a global variable - then also putting it in the my.cnf OR can you make the change in the my.cnf and just restart the server?