holland-discuss team mailing list archive

[Andrew Garner <muzazzi@xxxxxxxxx>]

Yes, I can reproduce this.   I agree the slave should not be stopped
at all - this is certainly not the intended behavior in dry-run mode.

We had been previously stubbing out the stop-slave routine so we
pretended to run it (and tested the code path) but never actually sent
a query to mysql.  We had changed stop-slave behavior slightly to only
stop the SQL_THREAD in 1.0.6 and this changed the method behavior that
broke the stubbing process - but the restart slave remained stubbed
out.  This means a stop slave (incorrectly) runs in dry-run but start
slave does not.

I am pushing a fix for this now.  This will be in our github repo
shortly and updated in our nightly snapshot later.

~Andy

On Mon, Jan 24, 2011 at 11:31 AM, Brandon Metcalf <bwmetcalf@xxxxxxxxx> wrote:
> If I run a dry-run with the mysqldump plugin and stop-slave set to
> yes, the slave is stopped but not restarted.  Anyone else experiencing
> this?  It seems like the slave shouldn't be stopped at all.
>
> _______________________________________________
> Mailing list: https://launchpad.net/~holland-discuss
> Post to     : holland-discuss@xxxxxxxxxxxxxxxxxxx
> Unsubscribe : https://launchpad.net/~holland-discuss
> More help   : https://help.launchpad.net/ListHelp
>