Original: https: //www.2cto.com/database/201708/674665.html
There are two ideas of data synchronization section, 1.master send only needed; 2.slave receive only desired.
master side:
binlog-do-db binary logging database (multi-database comma separated)
binlog-ignore-db ignore the binary log database (multi-database comma separated)
for example:
1) binlog-do-db = YYY database to be synchronized, not including the sync. (Do not add this line represents synchronize all)
Here only the main library synchronization test1, test2 library.
2) binlog-ignore-db = mysql This is not recorded binlog, to achieve from a library of sync mysql database to ensure that their rights
binlog-ignore-db=performance_schema
binlog-ignore-db=information_schema
Here to synchronize ignored test1, test2 library from the library
slave端
replicate-do-db is set to be copied database (multi-database using a comma separated)
replicate-ignore-db database replication is set to be ignored (multi-database using comma separated)
replicate-do-table setting table to be copied
replicate-ignore-table set replicated tables to be ignored
replicate-wild-do-table replication-do-table with the same function, but wildcards
replicate-wild-ignore-table like replication-ignore-table functions, but may add wildcards
The above comparison, replicate here on a good understanding, simply following points.
E.g:
Ignore test3 database copy from the library, but it should be noted that, in fact, from relaylog library from the relevant log on test3, but nothing from the library is not used.
Add two configuration wildcards
replicate-wild-do-table = db_name.% which library which table only replication
replicate-wild-ignore-table = mysql.% which table which library to ignore
【Finish】