[Pacemaker] Split Site 2-way clusters

Lars Ellenberg lars.ellenberg at linbit.com
Mon Jan 18 08:47:34 EST 2010


On Mon, Jan 18, 2010 at 11:14:58AM +0100, Andrew Beekhof wrote:
> > NodeX(Successfully) taking on data from clients while in
> > quorumless-freeze-still-providing-service, then discarding its hitherto
> > collected client data when realizing other node has quorum and discarding
> > own data isn’t good.
> 
> Agreed - freeze isn't an option if you're doing master/master.

no-quorum=freeze alone is not sufficient when doing master/slave, either:
if the current master risks to be blown away later,
you lose all changes from replication link loss to being shot.

So you have to make sure there will be no changes between those two
events, you need to also freeze IO on the DRBD Primary.  The fence-peer
handler script hook, and the DRBD fencing policy are what can be used
for this.

-- 
: Lars Ellenberg
: LINBIT | Your Way to High Availability
: DRBD/HA support and consulting http://www.linbit.com

DRBD® and LINBIT® are registered trademarks of LINBIT, Austria.




More information about the Pacemaker mailing list