sometimes MySQL replication may cause to broken for some reasons. Most of the time it’s either application errors due to poor table normalization or any direct table edit which done by the programmer.
One of the servers has got frequent this type of errors ( it is not replication errors) which shows poor slave health. So we need to analyse the problematic table.
What I used to do, run a table row count for both master and slave. Then skip this error by executing following commands
This is skip the problematic binary log and try to execute next log position and then continue. So that you will see the following errors on screen.
I’d also seen adding slave-skip_errors in “/etc/my.cnf” and they think it’s safe. But don’t expect one of these options are safe which will reduce the replication integrity.
few known error log numbers are shown below,
This would required MySQL restart so it is not advisable unless you know what exactly you are doing with.