[DRBD-user] activating and installing one node without another

Bart Coninckx bart.coninckx at telenet.be
Mon Sep 21 12:56:33 CEST 2009

Note: "permalinks" may not be as permanent as we would like,
direct links of old sources may well be a few messages off.


On Monday 21 September 2009 12:49, Mark Watts wrote:
> On Mon, 2009-09-21 at 11:12 +0200, Bart Coninckx wrote:
> > Hi,
> >
> > I need to migrate an existing server to a DRBD-Heartbeat cluster. The
> > existing server unfortunately hasn't got enough diskspace available to
> > accomodate the DRBD partition, so I need to reinstall it. My plan is to
> > install the second node, configure DRBD on it with an absent second node
> > and then completely reinstall the first node. This would require me to
> > bring "up" the partition on the DRBD device without the other DRBD device
> > on the other server being available until later. Is this possible?
>
> Yes.
>
> Do all the configuration as though there were two nodes, but set
> "wfc-timeout 0;" to stop that node waiting for a partner.
> Once you promote it to primary (probably via a split heartbeat cluster)
> you should be good to go. (/proc/drbd should show Primary/Unknown)
>
> When you bring the second node up (make sure the partition/device size
> is the same!) and you've done the
> 	"drbdadm create-md $resource"
> you should be able to
> 	"drbdadm -- --overwrite-data-of-peer primary $resource"
> on the primary and watch them sync.
>
> Mark.

Mark,

excellent, will try it this way!

B.



More information about the drbd-user mailing list