We have MySQL 5.5 multi-master replication configured on two servers. It worked fine for a long time. However, since some network problems have started to occur in our data center, we noticed an unexpected behaviour in this replication. When network problems* occur, we have both "Slave_IO_Running" & "Slave_SQL_Running" in "yes" state and "Seconds_Behind_Master" is 0 but the actual replication doesn't work. Why we think so? Positions from slave (relay log) doesn't change, slave operations stop to perform on master (no SQL queires from slave are executed on master). After doing simple stop slave & start slave this problem goes away.
We've updated MySQL to the last version (5.5.40) but nothing changed.
Is it a bug in Slave_IO_Running state? What can be done for more detailed debugging and fixing the problem?
* Unfortunately, we aren't able to understand the nature of these network problems for now.
We've updated MySQL to the last version (5.5.40) but nothing changed.
Is it a bug in Slave_IO_Running state? What can be done for more detailed debugging and fixing the problem?
* Unfortunately, we aren't able to understand the nature of these network problems for now.