Issue

• A cluster node logged the message "FAILED TO RECEIVE" and other node had lots of "Retransmit List" messages. Then the cluster nodes go into split-brain.
• When starting cman on one node after a reboot, it hangs a long time on 'starting fencing' and never seems to become a member. If I check the logs, I see several cpg_join errors followed by "FAILED TO RECEIVE"

  Mar 10 15:26:13 node2 openais[11158]: [CLM  ] CLM CONFIGURATION CHANGE 
  Mar 10 15:26:13 node2 openais[11158]: [CLM  ] New Configuration: 
  Mar 10 15:26:13 node2 openais[11158]: [CLM  ]     r(0) ip(192.168.10.10)  
  Mar 10 15:26:13 node2 openais[11158]: [CLM  ]     r(0) ip(192.168.10.11)  
  Mar 10 15:26:13 node2 openais[11158]: [CLM  ] Members Left: 
  Mar 10 15:26:13 node2 openais[11158]: [CLM  ] Members Joined: 
  Mar 10 15:26:13 node2 openais[11158]: [CLM  ]     r(0) ip(192.168.10.10)  
  Mar 10 15:26:13 node2 openais[11158]: [SYNC ] This node is within the primary component and will provide service. 
  Mar 10 15:26:13 node2 openais[11158]: [TOTEM] entering OPERATIONAL state. 
  Mar 10 15:26:13 node2 openais[11158]: [CLM  ] got nodejoin message 192.168.10.10
  Mar 10 15:26:13 node2 openais[11158]: [CLM  ] got nodejoin message 192.168.10.11 
  Mar 10 15:26:25 node2 groupd[11171]: cpg_join error retrying
  Mar 10 15:27:05 node2 last message repeated 4 times
  Mar 10 15:28:15 node2 last message repeated 7 times
  Mar 10 15:29:25 node2 last message repeated 7 times
  Mar 10 15:30:25 node2 last message repeated 6 times
  Mar 10 15:30:35 node2 groupd[11171]: cpg_join error retrying