Christopher Spring

I recently noticed that the database backups of an app. that i maintain were the same size for a couple of days.

Now, all the backups are done from a MySQL slave database, so it didn’t take too long for me to work out that my slave had died; running “show slave status\G” yielded something like this:

    ....
    Slave_IO_Running: Yes
    Slave_SQL_Running: No
    ....
    Seconds_Behind_Master: NULL

The trick here ended up being the following short little routine (after ascertaining that the error was not that serious - check syslog etc).

While in the mysql prompt, issue the following

    STOP SLAVE;
    SET GLOBAL SQL_SLAVE_SKIP_COUNTER = 1;
    START SLAVE;
    SHOW SLAVE STATUS\G

And if everything went according to plan, running “show slave status\G” should now yield:

    ....
    Slave_IO_Running: Yes
    Slave_SQL_Running: Yes
    ....
    Seconds_Behind_Master: [a few seconds]