> Additionally, what would I need to do to allow the cluster to failover if eth0 is detected to have gone down on one server? Is that another resource? Look at pingd: http://wiki.linux-ha.org/pingd Regards Dominik