Description of problem:
=======================
Was moving multiple temporary files to the same destination concurrently from 8 clients. The script ran for almost 3 hrs without any issues. After that started seeing continuous EEXIST errors from all the client mount points. 
When stopped the script from all the clients, there are many temp file because of the rename failures and also we are seeing two files with same name.

-rw-r--r--. 1 root root 10 Jul 31 15:23 test
-rw-r--r--. 1 root root 10 Jul 31 15:23 test
-rw-r--r--. 1 root root 10 Jul 31 15:21 test00264710-74c8-4d1b-b446-b4ee51a6b89f
-rw-r--r--. 1 root root 10 Jul 31 15:23 test00743e28-e816-4dde-a80b-0ffa50d6ab86
-rw-r--r--. 1 root root 10 Jul 31 15:21 test00dcbe58-12ee-4e6f-8c73-ecbd84798e72
-rw-r--r--. 1 root root 10 Jul 31 15:22 test010dd93c-29b0-4a70-bb87-e88ed519105d
-rw-r--r--. 1 root root 10 Jul 31 15:23 test012a6482-825e-471b-b27a-5ecc7c77a971
.... many

Version-Release number of selected component (if applicable):
3.12.2-14.el7rhgs.x86_64

How reproducible:
always

Steps to Reproduce:
====================
1) Create a distributed-replicated volume and start it.
2) FUSE mount it on multiple clients.
3) From all clients execute,

"while true; do uuid="`uuidgen`"; echo "some data" > "test$uuid"; mv "test$uuid" "test" -f; done"

4) Wait until you hit EEXIST errors. (I hit it after running for almost 3 hrs)

Actual results:
================
Moving multiple temporary files to the same destination concurrently throws EEXIST error 

Expected results:
==================
Moving multiple temporary files to the same destination concurrently should not throw any error. 

Additional info:
================
Sos and gluster-health-check reports will be shared.