Description of problem:
=======================
On ubuntu cluster purge-cluster fails with error - stderr: grep: write error: 



Version-Release number of selected component (if applicable):
=============================================================



How reproducible:
================
always


Steps to Reproduce:
===================
1. created ubuntu cluster with one MON and 3 OSD node
2. Did some I/O using rgw and rados
3. purge cluster using below command:-

oot@magna044 ceph-ansible]#  ansible-playbook purge-cluster.yml -vv -i  /etc/ansible/ubuntu  --extra-vars '{"ceph_stable": true, "ceph_stable_rh_storage_cdn_install": true , "ceph_stable_rh_storage": true, "monitor_interface": "eth0", "journal_collocation": true, "devices": ["/dev/sdb", "/dev/sdc", "/dev/sdd"], "journal_size": 100, "public_network": "x.x.x.x/21","cephx": true, "fetch_directory": "~/ubu",  "calamari":true, "radosgw_civetweb_port": "8080" }' -u ubuntu

Actual results:
===============

TASK: [check for anything running ceph] *************************************** 
<magna028> REMOTE_MODULE command -  #USE_SHELL
failed: [magna028] => {"changed": true, "cmd": "ps awux | grep -v grep | grep -q -- ceph-", "delta": "0:00:00.024376", "end": "2016-07-16 14:41:32.131136", "failed": true, "failed_when_result": true, "rc": 0, "start": "2016-07-16 14:41:32.106760", "stdout_lines": [], "warnings": []}
stderr: grep: write error: 

FATAL: all hosts have already failed -- aborting



Additional info: