Why is the Gluster Client Process "glusterfs" Crashing at file descriptors operations?

Issue

• The glusterfs process in a Gluster client is crashing with the following stacktrace:

  gdb /usr/sbin/glusterfs core.file
  
  [...]
  
  Thread 1 (Thread 0x7f06d83f4700 (LWP 78393)):
  #0  __GI___pthread_mutex_lock (mutex=0x18) at ../nptl /pthread_mutex_lock.c:65
  #1  0x00007f06d9afc5a5 in ob_fd_free     (ob_fd=ob_fd@entry=0x7f06a001dec0)
      at open-behind.c:198
  #2  0x00007f06d9afcec9 in ob_inode_wake  (this=this@entry=0x7f06d4014a40,
      ob_fds=ob_fds@entry=0x7f06d83f2e80) at open-behind.c:355
  #3  0x00007f06d9afd03c in open_all_pending_fds_and_resume (
      this=this@entry=0x7f06d4014a40, inode=<optimized out>,  stub=0x7f06a000c3c8)
      at open-behind.c:442
  #4  0x00007f06d9afd483 in ob_rename (frame=0x7f06a001e168,
      this=0x7f06d4014a40, src=<optimized out>, dst=0x7f06c8098f00,
      xdata=<optimized out>) at open-behind.c:1035
  #5  0x00007f06e8ec7441 in default_icreate_cbk (frame=0x18,
      cookie=0x7f06d400f0b0, this=0x7f06c80818e0, op_ret=0,      op_errno=0,
      inode=0x0, buf=0x7f0600000000, xdata=0x0) at defaults.c:1709
  #6  0x00007f06d96cd8e8 in mdc_rename (frame=0x7f06a00153d8,
      this=0x7f06d40180d0, oldloc=0x7f06c8098ec0,   newloc=0x7f06c8098f00,
      xdata=0x0) at md-cache.c:1833
  #7  0x00007f06e8edde55 in default_icreate_cbk_resume  (frame=0x7f06d4019bb0,
      cookie=<optimized out>, this=<optimized out>, op_ret=-1610525736,
      op_errno=-647178576, inode=0x7f06c8098f00,    buf=0x7f06d4019bb0,
      xdata=0x7f06e8e5dd71 <call_resume_wind+1905>) at defaults.c:1084
  
  [...]
  

• How to prevent this type of crashes?