Why brick report its port as "N/A" in the output of 'gluster volume status' command in Red Hat Gluster Storage 3.0 and simultaneously RHEV-M shows brick down ?

Issue

• Brick status invalid in RHEV-M UI

• Running the gluster volume status command, a brick reports its port as N/A, but the volume otherwise seems operational.

Status of volume: rhs_vol
Gluster process                                                            TCP Port  RDMA Port  Online  Pid
------------------------------------------------------------------------------
Brick rhs-storage.redhat.com:/rhs/brick                       N/A            N/A        Y       10684

~Snippet from vdsm log on hypervisor or RHS nodes,

Detector thread::DEBUG::2015-07-07 04:01:03,400::protocoldetector::207::vds.MultiProtocolAcceptor::(_process_handshake) Error during handshake: unexpected eof
Detector thread::DEBUG::2015-07-07 04:01:03,400::protocoldetector::207::vds.MultiProtocolAcceptor::(_process_handshake) Error during handshake: unexpected

• Experienced this issue after upgrade to RHEV-M 3.5 (latest available). One RHS brick is marked down in UI although gluster status gathered from CLI is ok

# gstatus -a -b
      Product: Community          Capacity:  29.10 TiB(raw bricks)     
      Status: HEALTHY                        3.80 TiB(raw used)
   Glusterfs: 3.6.0.53                      14.50 TiB(usable from volumes)
  OverCommit: No                Snapshots:   0

   Nodes    :  5/ 5     Volumes:  7 Up
   Self Heal:  4/ 4               0 Up(Degraded)
   Bricks   : 28/28               0 Up(Partial)
   Clients  :     5                       0 Down