Description of problem:
=======================
After rebalance is triggered on the volume, on few files I am seeing ENOSPC errors in rebalance logs though there is enough space left on the bricks.

2018-02-20 05:31:01.051156] E [MSGID: 109023] [dht-rebalance.c:2749:gf_defrag_migrate_single_file] 0-distrepx3-dht: migrate-data on /linux-4.6.4/Documentation/ABI/testing/configfs-usb-gadget-serial failed: [No space left on device]
[2018-02-20 05:31:01.139027] E [MSGID: 109023] [dht-rebalance.c:2749:gf_defrag_migrate_single_file] 0-distrepx3-dht: migrate-data on /linux-4.6.4/Documentation/ABI/testing/debugfs-pktcdvd failed: [No space left on device]
[2018-02-20 05:31:01.769876] E [MSGID: 109023] [dht-rebalance.c:2749:gf_defrag_migrate_single_file] 0-distrepx3-dht: migrate-data on /linux-4.6.4/Documentation/ABI/testing/sysfs-bus-iio-meas-spec failed: [No space left on device]
[2018-02-20 05:31:03.973459] E [MSGID: 109023] [dht-rebalance.c:2749:gf_defrag_migrate_single_file] 0-distrepx3-dht: migrate-data on /linux-4.6.4/Documentation/ABI/testing/sysfs-ibft failed: [No space left on device]
[2018-02-20 05:31:09.066247] E [MSGID: 109023] [dht-rebalance.c:2749:gf_defrag_migrate_single_file] 0-distrepx3-dht: migrate-data on /linux-4.6.4/Documentation/DocBook/media/v4l/vidioc-g-fbuf.xml failed: [No space left on device]
[2018-02-20 05:31:13.597532] E [MSGID: 109023] [dht-rebalance.c:2749:gf_defrag_migrate_single_file] 0-distrepx3-dht: migrate-data on /linux-4.6.4/Documentation/RCU/trace.txt failed: [No space left on device]
[2018-02-20 05:31:24.303807] E [MSGID: 109023] [dht-rebalance.c:2749:gf_defrag_migrate_single_file] 0-distrepx3-dht: migrate-data on /linux-4.6.4/Documentation/cgroup-v1/cgroups.txt failed: [No space left on device]
[2018-02-20 05:31:24.992058] E [MSGID: 109023] [dht-rebalance.c:2749:gf_defrag_migrate_single_file] 0-distrepx3-dht: migrate-data on /linux-4.6.4/Documentation/connector/connector.txt failed: [No space left on device]
[2018-02-20 05:31:30.169573] E [MSGID: 109023] [dht-rebalance.c:2749:gf_defrag_migrate_single_file] 0-distrepx3-dht: migrate-data on /linux-4.6.4/Documentation/devicetree/bindings/arm/calxeda/l2ecc.txt failed: [No space left on device]
[2018-02-20 05:31:33.802256] E [MSGID: 109023] [dht-rebalance.c:2749:gf_defrag_migrate_single_file] 0-distrepx3-dht: migrate-data on /linux-4.6.4/Documentation/devicetree/bindings/arm/omap/crossbar.txt failed: [No space left on device]
[2018-02-20 05:31:39.996225] E [MSGID: 109023] [dht-rebalance.c:2749:gf_defrag_migrate_single_file] 0-distrepx3-dht: migrate-data on /linux-4.6.4/Documentation/devicetree/bindings/clock/ti/mux.txt failed: [No space left on device]
[2018-02-20 05:31:40.264191] E [MSGID: 109023] [dht-rebalance.c:2749:gf_defrag_migrate_single_file] 0-distrepx3-dht: migrate-data on /linux-4.6.4/Documentation/devicetree/bindings/clock/clk-palmas-clk32kg-clocks.txt failed: [No space left on device]

Grep output for a file in rebalance logs:

# grep -i /linux-4.6.4/Documentation/hsi.txt /var/log/glusterfs/distrepx3-rebalance.log 
[2018-02-20 05:33:04.980564] I [dht-rebalance.c:1513:dht_migrate_file] 0-distrepx3-dht: /linux-4.6.4/Documentation/hsi.txt: attempting to move from distrepx3-replicate-0 to distrepx3-replicate-2
[2018-02-20 05:33:05.096191] W [MSGID: 109023] [dht-rebalance.c:962:__dht_check_free_space] 0-distrepx3-dht: data movement of file {blocks:6 name:(/linux-4.6.4/Documentation/hsi.txt) } would result in dst node (distrepx3-replicate-2:41451088) having lower disk space then the source node (distrepx3-replicate-0:41453744).Skipping file.
[2018-02-20 05:33:05.127613] I [MSGID: 109126] [dht-rebalance.c:2714:gf_defrag_migrate_single_file] 0-distrepx3-dht: File migration skipped for /linux-4.6.4/Documentation/hsi.txt.
[2018-02-20 05:33:05.127739] E [MSGID: 109023] [dht-rebalance.c:2749:gf_defrag_migrate_single_file] 0-distrepx3-dht: migrate-data on /linux-4.6.4/Documentation/hsi.txt failed: [No space left on device]


Version-Release number of selected component (if applicable):
3.12.2-4.el7rhgs.x86_64

How reproducible:
1/1

Steps to Reproduce:
===================
1) Create a x3 volume and start it.
2) FUSE mount it on multiple clients.
3) Run linux kernel untar from two clients .
4) While IO is in-progress, add bricks to the volume and start rebalance without force.

Actual results:
===============
Seeing ENOSPC errors on few files in rebalance logs though there is enough space left on the bricks.

Expected results:
=================
NO errors in rebalance logs.