On Tue, Jul 07, 2009 at 01:50:17PM +0200, Guido Trotter wrote:
> On Tue, Jul 7, 2009 at 12:54 PM, Iustin Pop<[email protected]> wrote:
> 
> > s/second node/original master/ ?
> >
> 
> Will fix
> 
> >> +        down). After a failover performed this way the master daemon will 
> >> most
> >> +        probably not start, and you will need to start it manually 
> >> passing the
> >> +        --no-voting option to ganeti-masterd as well.
> >
> > Hmm, I think that we should modify the start master call over RPC to
> > allow forcing the start, so that failover --no-voting is enough.
> >
> 
> This also needs a change in ganeti-masterd to make --no-voting
> (optionally) non intreractive.
> Are we fine with that?

Yes, but a different option --yes-really-not-voting (or --no-voting
--yes-i-really-mean-it).

> >> Be careful because the
> >> +        second node will still believe to be the master, so when it comes 
> >> up
> >> +        you'll need to start just ganeti-noded there, and perform a 
> >> gnt-cluster
> >> +        redist-conf on the new master to make the cluster consistent 
> >> again.
> >> +      </para>
> >
> > IIRC, the masterd won't start on the original master since the vote from
> > the current master will conflict. We should check this.
> >
> 
> Yes, you're right! Testing now, and will fix the manpage about this if
> everything works as intended.

Thanks!

iustin

Reply via email to