Update time sync in devops script

[Oleksiy Molchanov]

Please update devops scripts with the following:

1. Change
  NTPD=$(find /etc/init.d/ -regex '/etc/init.d/ntp.?'); $NTPD stop; killall ntpd; ntpd -qg && $NTPD start
  to
  crm_resource --list | grep -q "p_ntp" && { crm resource stop p_ntp; ntpd -qg && crm resource start p_ntp ; } || { NTPD=$(find /etc/init.d/ -regex \'/etc/init.d/ntp.?\'); $NTPD stop; killall ntpd; ntpd -qg && $NTPD start ; }

2. The env reverting mechanism must work in such way, that on first stage all the controllers are supposed to be reverted, then rest of the nodes.
3. Replace all ntpdate commands mentioned in scripts to ntpd -qg

[Oleksiy Molchanov (omolchanov)]

please consider \'/etc/init.d/ntp.?\' = '/etc/init.d/ntp.?'

[Dennis Dmitriev (ddmitriev)]

1. 'crm_resource --list' must not be used in fuel-devops. Not all environments are deployed in HA mode. Not all environments even have operation system installed on slaves , for example snapshots ready_with_3_slaves, ready_with_5_slaves etc.

2. Before making a snapshot, all environments are paused. There is no need to follow the specific order during revert because all nodes are resumed simultaneously.