One of the controller has no resources or can't join cluster in OpenStack

Issue

• After an outage, it happens that a controller has no resources, or can't join the cluster

[root@controller1 ~]# pcs status
Cluster name: tripleo_cluster
WARNING: no stonith devices and stonith-enabled is not false
Stack: corosync
Current DC: controller1 (version 1.1.15-11.el7_3.4-e174ec8) - partition WITHOUT quorum
Last updated: Tue May 29 16:15:55 2018          Last change: Tue May 29 16:14:19 2018 by hacluster via crmd on controller1

3 nodes and 0 resources configured

Node controller0: UNCLEAN (offline)
Node controller2: UNCLEAN (offline)
Online: [ controller1 ]

No resources


Daemon Status:
  corosync: active/enabled
  pacemaker: active/enabled
  pcsd: active/enabled