MogileFS limits...
Nathan Schmidt
nschmidt at gmail.com
Sun Apr 13 04:28:28 UTC 2008
Heh, yeah, we run master-master but I wasn't thrilled with the dance.
We'd be replacing the hardware too, so it'd be more like A-B slaving
to C, which is a new InnoDB box, and then A-C, and then A-C slaving to
D, which is the other new InnoDB box, and then C-D master-master (plus
a slave backup on E...)
Sounds like a hassle to me, which is why we've hired people who are
better at ops. :)
-n
On Sat, Apr 12, 2008 at 7:42 PM, Ask Bjørn Hansen <ask at develooper.com> wrote:
>
> On Apr 12, 2008, at 16:48, Nathan Schmidt wrote:
>
>
> > Thanks for the tips! We've got a project underway already to move to
> > InnoDB but it's a bit of a hassle (particularly since we really really
> > don't like going offline for maintenance).
> >
>
>
> Don't you run master-master?
>
> Stop replication from A-B, run database traffic on "B", do schema changes
> on "A", let replication (B-A) catch up, move traffic to "A", restart A-B
> replication.
>
> Viola. No downtime.
>
>
>
>
> - ask
>
> --
> http://www.askbjoernhansen.com/
>
>
>
--
New office! 1825 South Grant Street, Suite 850 San Mateo 94402
New home! 21677 Rainbow Drive Cupertino 95014
New phone! 415.420.1647
More information about the mogilefs
mailing list