JFL wrote:
> I need to restore 20 gigabytes of binary logfiles.
> What should I do in order to get the job done as quickly as possible?
>
> I am considering this:
> mysqlbinlog --database=mydb logfiles | mysql -f
There is a faster way, but it is rather tricky.
4.0 slave can be tricked into thinking that those binary logs are in fact the
relay logs that it gathered from the master. So we trick the server into
thinking it is a slave of the server that produced them. You just need to pick a
server-id that is not the same as the one that produced them. Then if you list
the logs in in the relay log info file in the correct order, hand-craft
relay-log.info file to point at the first one), then start the slave with
skip-slave-start, and then just start the SQL thread manually (SLAVE START
SQL_THREAD), it will process them until it is done. Periodically run SHOW SLAVE
STATUS to see if it got to the end of the last log.
--
Sasha Pachev
Create online surveys at http://www.surveyz.com/