After rebooting a hyperconverged RHV host, another host reports gluster peer in cluster disconnected state for the rebooted host

Issue

• After rebooting a hyperconverged RHV-H host gracefully or ungracefully the host boots up, but a different host in the same cluster reports the gluster peer state of the rebooted host as Disconnected.

  # gluster peer status
  Number of Peers: 5
  
  Hostname: 172.29.60.11
  Uuid: 3b84c072-af3f-4d77-bc03-7a19f51e2c14
  State: Peer in Cluster (Connected)
  
  Hostname: 172.29.60.13
  Uuid: a80199cb-fece-46ef-9d78-67205ea57525
  State: Peer in Cluster (Connected)
  
  Hostname: 172.29.60.14
  Uuid: 0726b43b-b401-4a71-884a-f4402912ced4
  State: Peer in Cluster (Connected)
  
  Hostname: 172.29.60.15
  Uuid: 20ea94db-53a5-4d7d-9b40-4a845c6be49e
  State: Peer in Cluster (Disconnected)
  
  Hostname: 172.29.60.16
  Uuid: 193d60e4-0417-4241-a5b3-f34fd40ab5c5
  State: Peer in Cluster (Connected)
  

• Restarting the glusterd service on the Disconnected peer temporarily clears the issue and all peers display Connected state until the next reboot.