RHEL 6 High Availbility Cluster with cman fails to initialize qdisk at boot time when the quorum disk is an FCoE device

Issue

• When setting an FCoE device as a qdisk, it fails to initialize at boot time causing cluster startup to fail
• Unable to detect FCoE devices at boot time before the FCoE init script deploys

Aug 12 19:13:40 node1 corosync[2779]:   [MAIN  ] Corosync Cluster Engine ('1.4.7'): started and ready to provide service.
Aug 12 19:13:40 node1 corosync[2779]:   [MAIN  ] Corosync built-in features: nss dbus rdma snmp
Aug 12 19:13:40 node1 corosync[2779]:   [MAIN  ] Successfully read config from /etc/cluster/cluster.conf
Aug 12 19:13:40 node1 corosync[2779]:   [MAIN  ] Successfully parsed cman config
[....]
Aug 12 19:13:45 node1 qdiskd[2842]: Unable to match label 'quorum' to any device
Aug 12 19:13:47 node1 corosync[2779]:   [SERV  ] Unloading all Corosync service engines.
Aug 12 19:13:47 node1 corosync[2779]:   [SERV  ] Service engine unloaded: corosync extended virtual synchrony service
[...]
Aug 12 19:13:48 node1 kernel: cnic: Broadcom NetXtreme II CNIC Driver cnic v2.5.20 (March 14, 2014)
Aug 12 19:13:48 node1 kernel: bnx2x 0000:06:00.2: eth2: Added CNIC device
Aug 12 19:13:48 node1 kernel: bnx2x 0000:06:00.3: eth3: Added CNIC device
Aug 12 19:13:48 node1 kernel: bnx2fc: Broadcom NetXtreme II FCoE Driver bnx2fc v2.4.2 (Dec 11, 2013)
Aug 12 19:13:48 node1 kernel: bnx2fc: [06]: FCOE_INIT passed
Aug 12 19:13:48 node1 kernel: bnx2fc: [06]: FCOE_INIT passed