Description of problem:
===========================
When running ping_pong on a file ("ping_pong -rw testfile N") from multiple fuse mounts,  the "data increment" value is not equal the number of nodes currently running the ping_pong test. The "data increment" value is not getting incremented when a new instance of ping_pong is started. 

Version-Release number of selected component (if applicable):
==============================================================
glusterfs 3.4.0.27rhs built on Aug 29 2013 09:57:33

How reproducible:
====================
Often

Steps to Reproduce:
====================
1. Create a replicate volume ( 1 x 2 ) 

2. Set the following volume options:
write-behind off
eager-lock off
storage.batch-fsync-delay-usec 0

3. Start the volume.

4. Create 4 fuse mounts from a client node. 

5. Start an instance of ping_pong on mount1 : "ping_pong -rw testfile 6" . 
"data increment" value should be "1" and it should also print a lock rate once per second

Actaul result : 

Mount1:- 
++++++++++++++++++
root@darrel [Aug-29-2013-17:47:59] >ping_pong -rw ./testfile 6
data increment = 1
   166 locks/sec

6. Start another instance of ping_pong on mount2 :"ping_pong -rw testfile 6"
"data increment" value should be "2" and it should also print a lock rate once per second

Actual results:

Mount1:- 
++++++++++++++++++
root@darrel [Aug-29-2013-17:47:59] >ping_pong -rw ./testfile 6
data increment = 1
   137 locks/sec
    85 locks/sec

Mount2:-
++++++++++++++++++
root@darrel [Aug-29-2013-17:55:33] >ping_pong -rw ./testfile 6
data increment = 1
   134 locks/sec

Expected results:
====================
"data increment" value should be incremented and should be equal to number of nodes currently running the ping_pong test