MySQL Manual
4 Database Administration
- 4.10 Replication in MySQL
- 4.10.7 SQL Commands Related to Replication
- 4.10.7.1
START SLAVE(slave) - 4.10.7.2
STOP SLAVE(slave) - 4.10.7.3
SET SQL_LOG_BIN=0|1(master) - 4.10.7.4
SET GLOBAL SQL_SLAVE_SKIP_COUNTER=n(slave) - 4.10.7.5
RESET MASTER(master) - 4.10.7.6
RESET SLAVE(slave) - 4.10.7.7
LOAD TABLE tblname FROM MASTER(slave) - 4.10.7.8
LOAD DATA FROM MASTER(slave) - 4.10.7.9
CHANGE MASTER TO master_def_list(slave) - 4.10.7.10
MASTER_POS_WAIT()(slave) - 4.10.7.11
SHOW MASTER STATUS(master) - 4.10.7.12
SHOW SLAVE HOSTS(master) - 4.10.7.13
SHOW SLAVE STATUS(slave) - 4.10.7.14
SHOW MASTER LOGS(master) - 4.10.7.15
SHOW BINLOG EVENTS(master) - 4.10.7.16
PURGE MASTER LOGS(master)
- 4.10.7.1
4.10.7.10 MASTER_POS_WAIT() (slave)
This is not a command but a function, used to ensure that the slave has reached (read and executed up to) a given position in the master's binlog; see section 6.3.6.2 Miscellaneous Functions for a full description.
User Comments
| Posted by Nathan Tanner on Saturday August 9 2003, @10:13am | [Delete] [Edit] |
NOTE: If you have a cyclical replication setup (ie. more then two servers in a replication chain) do *NOT* rely on 'SHOW MASTER STATUS' on the slave's MASTER for values to use in the master_pos_wait function. This is because if the slave had just finished executing a binlog event that didn't originally come from its own MASTER, the master_log_pos value to be used in this function won't work. I feel this is a bug, but it is apparently designed to work this way. Don't ask me why.
