Hi all,
I was wodnering if someone has a better way of re-sync'ing a slave when
replication fails.
Here's my setup:
1 - MASTER
5 - SLAVES
Now sometimes, one of the slaves will fail. Usually when it's the sql that stops
running, I can fix it and get it back. However sometimes, I also get the "master
binlog corruption" error and haven't found a way to fix that.
What I do is:
- stop mysql on the master and all the slaves
- remove the server-bin files on the master
- remove the mysql data dir on each slave
- copy the master's datadir to each slave
- restart mysql everywhere and everything is fine
Now while this method works, it is extremely time-consuming and I was wondering
if anyone knows a better solution to only fix the affected slave??
I have tried stopping just the affected slave and one other good slave, taking
its datadir and renaming the bin files to the affected server's name. That
method doesn't seem to work though.
Thanks in advance!
