Copy problems and solutions
Data corruption or loss of error
For a variety of reasons, MySQL replication is not well from a server crash, power failure, disk corruption, memory or network error recovery. Almost certainly need to be copied from the beginning to restart some point encounter these problems.
Most skin problems due to non-normal shutdown is caused by the absence of timely data to disk brush
The main and backup database storage engines use different libraries
Use different storage engines on standby database, it can sometimes be beneficial. However, in some scenarios, but use replication statement is based, if the standby database uses a different storage engine, it may result in a different execution results on the main library and library equipment, such as uncertainty statement uses different Zhu standby database more likely to cause problems during storage engine.
Non-replicated data dependency
If you have a standby database does not exist in the database or table in the main library, copying will be very easy to accidentally break, and vice versa. Suppose there is a library prepared does not exist on the primary database repository named scratch. If this occurs update the tables in the database on the primary database, standby database will attempt to interrupt the reproduction of these updates. Similarly, if you create a table on a standby database that already exists in the main library copy may also be interrupted.