Restoring a broken slave

[Dan Casey]

mysql 4.0.20
myIsam tables..

I have two slave setups. One slave is fine and up to date. The second one which also updates off the same master showed the following.


Slave_IO_Running: Yes
Slave_SQL_Running: No


I have had this issue before. In which case I run slave stop; slave start; and then it starts working again. This time around, Slave_SQL_Running remained as "NO"..
There was a duplicate entry (which im not sure how that happened here if it didn't on the other slave). Im guessing thats where it stopped?

Right now im reloading lastnights master snapshot.. just to give this slave a more current db.

My questions are these.

1. I know im going to have to make a new mysqldump of the master, and record its current position in order to get replication started again. Is there a way that I can do this without disrupting service on my web site.. Or am I going to have to expect some downtime while the master is out?

2. How did this happen, or better yet, what can or should i do in the future?


3. almost forget, but since i am completely making a new slave here, is it safe to rm -rf all the bin-log and relay-bin-logs on this slave

Thanks,