Each binary logging format has advantages and disadvantages. For most users, the mixed replication format should provide the best combination of data integrity and performance. If, however, you want to take advantage of the features specific to the statement-based or row-based replication format when performing certain tasks, you can use the information in this section, which provides a summary of their relative advantages and disadvantages, to determine which is best for your needs.
Proven technology.
Less data written to log files. When updates or deletes affect many rows, this results in much less storage space required for log files. This also means that taking and restoring from backups can be accomplished more quickly.
Log files contain all statements that made any changes, so they can be used to audit the database.
Statements that are unsafe for SBR. Not all statements which modify data (such as INSERT DELETE, UPDATE, and REPLACE statements) can be replicated using statement-based replication. Any nondeterministic behavior is difficult to replicate when using statement-based replication. Examples of such Data Modification Language (DML) statements include the following:
A statement that depends on a loadable function or stored program that is nondeterministic, since the value returned by such a function or stored program or depends on factors other than the parameters supplied to it. (Row-based replication, however, simply replicates the value returned by the function or stored program, so its effect on table rows and data is the same on both the source and replica.) See Section 17.5.1.16, “Replication of Invoked Features”, for more information.
DELETE and UPDATE statements that use a LIMIT clause without an ORDER BY are nondeterministic. See Section 17.5.1.18, “Replication and LIMIT”.
Locking read statements (SELECT ... FOR UPDATE and SELECT ... FOR SHARE) that use NOWAIT or SKIP LOCKED options. See Locking Read Concurrency with NOWAIT and SKIP LOCKED.
Deterministic loadable functions must be applied on the replicas.
Statements using any of the following functions cannot be replicated properly using statement-based replication:
SYSDATE() (unless both the source and the replica are started with the --sysdate-is-now option)
However, all other functions are replicated correctly using statement-based replication, including NOW() and so forth.
For more information, see Section 17.5.1.14, “Replication and System Functions”.
Statements that cannot be replicated correctly using statement-based replication are logged with a warning like the one shown here:
[Warning] Statement is not safe to log in statement format.
A similar warning is also issued to the client in such cases. The client can display it using SHOW WARNINGS.
INSERT ... SELECT requires a greater number of row-level locks than with row-based replication.
UPDATE statements that require a table scan (because no index is used in the WHERE clause) must lock a greater number of rows than with row-based replication.
For InnoDB: An INSERT statement that uses AUTO_INCREMENT blocks other nonconflicting INSERT statements.
For complex statements, the statement must be evaluated and executed on the replica before the rows are updated or inserted. With row-based replication, the replica only has to modify the affected rows, not execute the full statement.
If there is an error in evaluation on the replica, particularly when executing complex statements, statement-based replication may slowly increase the margin of error across the affected rows over time. See Section 17.5.1.29, “Replica Errors During Replication”.
Stored functions execute with the same NOW() value as the calling statement. However, this is not true of stored procedures.
Deterministic loadable functions must be applied on the replicas.
Table definitions must be (nearly) identical on source and replica. See Section 17.5.1.9, “Replication with Differing Table Definitions on Source and Replica”, for more information.
As of MySQL 8.0.22, DML operations that read data from MySQL grant tables (through a join list or subquery) but do not modify them are performed as non-locking reads on the MySQL grant tables and are therefore not safe for statement-based replication. For more information, see Grant Table Concurrency.
All changes can be replicated. This is the safest form of replication.
Statements that update the information in the mysql system schema, such as GRANT, REVOKE and the manipulation of triggers, stored routines (including stored procedures), and views, are all replicated to replicas using statement-based replication.
For statements such as CREATE TABLE ... SELECT, a CREATE statement is generated from the table definition and replicated using statement-based format, while the row insertions are replicated using row-based format.
Fewer row locks are required on the source, which thus achieves higher concurrency, for the following types of statements:
Fewer row locks are required on the replica for any INSERT, UPDATE, or DELETE statement.
RBR can generate more data that must be logged. To replicate a DML statement (such as an UPDATE or DELETE statement), statement-based replication writes only the statement to the binary log. By contrast, row-based replication writes each changed row to the binary log. If the statement changes many rows, row-based replication may write significantly more data to the binary log; this is true even for statements that are rolled back. This also means that making and restoring a backup can require more time. In addition, the binary log is locked for a longer time to write the data, which may cause concurrency problems. Use binlog_row_image=minimal to reduce the disadvantage considerably.
Deterministic loadable functions that generate large BLOB values take longer to replicate with row-based replication than with statement-based replication. This is because the BLOB column value is logged, rather than the statement generating the data.
You cannot see on the replica what statements were received from the source and executed. However, you can see what data was changed using mysqlbinlog with the options --base64-output=DECODE-ROWS and --verbose.
Alternatively, use the binlog_rows_query_log_events variable, which if enabled adds a Rows_query event with the statement to mysqlbinlog output when the -vv option is used.
For tables using the MyISAM storage engine, a stronger lock is required on the replica for INSERT statements when applying them as row-based events to the binary log than when applying them as statements. This means that concurrent inserts on MyISAM tables are not supported when using row-based replication.