pcs - two nodes - Split-brain problems

Question

I have a cluster (pacemaker, corosysnc and pcs) with two nodes, no-quorum-polic is ignored. With resources for a LAMP server and a DRBD data partition. I have split-brain problems if they don't turn on and off in a certain order.

I would like to be able to always force the master to the same node, and for the change to the other node to be manual. Or I'm not sure if setting quorum-polic to freeze would solve this problem.

Resources

sudo pcs status resources
  * virtual_ip  (ocf::heartbeat:IPaddr2):    Started ibt1-ip192
  * Clone Set: DrbdData-clone [DrbdData] (promotable):
    * Masters: [ ibt1-ip192 ]
    * Slaves: [ ibt2-ip192 ]
  * DrbdFS  (ocf::heartbeat:Filesystem):     Started ibt1-ip192
  * WebServer   (ocf::heartbeat:apache):     Started ibt1-ip192
  * Maria   (ocf::heartbeat:mysql):  Started ibt1-ip192

Constraints

sudo pcs constraint

Location Constraints:
Ordering Constraints:
  promote DrbdData-clone then start DrbdFS (kind:Mandatory)
  start DrbdFS then start virtual_ip (kind:Mandatory)
  start virtual_ip then start WebServer (kind:Mandatory)
  start DrbdFS then start Maria (kind:Mandatory)
Colocation Constraints:
  DrbdFS with DrbdData-clone (score:INFINITY) (with-rsc-role:Master)
  virtual_ip with DrbdFS (score:INFINITY)
  WebServer with virtual_ip (score:INFINITY)
  Maria with DrbdFS (score:INFINITY)
Ticket Constraints:

Any idea how I can do it? I'm a bit lost. Thank you very much.