I'm running traditional master/slave replication on MySQL 5.1.52, my master is set for statement based replication. A few tables are missing rows on the slave (differing row counts), these tables are insert only on the master using "load data infile ...", with auto-increment Pks. It is strange that the missing rows on the slave have gaps in the PK values where the rows should be. Slave is marked read-only so there are no accidental deletes.
|
|
The traditional way would be to use pt-table-checksum and pt-table-sync I like doing things a little different. I immediately run Here is the --sync-to-master option
Here is the --print option
This, if you do something like this:
The file With regard to using LOAD DATA INFILE in Replication, @DTest answered this question about that. I further explained how mysql replicates LOAD DATA INFILE. |
|||||||
|
