Ed W <lists@wildgooses.com> wrote on 2010-Jul-27:
> I'm now investigating how to scale up from here and really how we can
> move from a simple backup of image/data files and move towards higher
> availability using a small pool of machines (possibly in different data
> centers) and more ability to move vservers between machines in the event
> of problems or downtime.
>
> At present I'm really just looking to achieve a goal of a backup machine
> which has a warm backup of our data and will be periodically (or better)
> synced with the main machine. Considering rsync or DRBD for this role
> (bit scared of DRBD though).
>
> We would like to have a plan to move up from there to a hot failover
> option involving multiple machines in the same data center and should a
> machine fail then we can move the services over to the new machine
> (target under 10 mins to failover). However, there seem to be
> significant issues in automating this and quite a bit of thought seems
> to be needed to fencing dead machines. Anyone got any experience to share?
DRBD + Heartbeat is very easy to use with vserver and it works beautifully.
Heartbeat and DRBD work well together; they include the necessary tools to fence
machines properly and insure that when failed machines come back up, they are
properly integrated back as the secondary node in the primary/secondary cluster.
We've been running all our services on two-node clusters like this for almost
three years (first using debian etch and now lenny as the host system) and never
had issues with split-brain or anything like that. And for lenny, at least,
everything you need is already in the repositories. If you've got two host
machines, you could have it up and running in a couple of hours.
I've been meaning to take all our internal documentation on our setup and put it
together in some coherent fashion for public consumption. Your post encourages
me to do that sooner rather than later! :-)
HTH
Jeff Jansen
Received on Wed Jul 28 07:26:30 2010