Comment 28 for bug 1385755

Reviewed: https://review.openstack.org/139053
Committed: https://git.openstack.org/cgit/stackforge/fuel-library/commit/?id=cb75bea0d0c9474c9d25502130f735399e8a94e5
Submitter: Jenkins
Branch: master

commit cb75bea0d0c9474c9d25502130f735399e8a94e5
Author: Bogdan Dobrelya <email address hidden>
Date: Thu Dec 4 14:37:09 2014 +0100

    Do additional checks in OCF only for action monitor.

    W/o this change, the *monitor() function of OCF scripts
    do additional AMQP connection checks unconditionally, every
    time it has been invoked. And that could cause *too much* netstat
    invocations for host OS to handle them w/o an issues.

    By design, we want these additional checks performed only when
    OCF script being invoked as a monitor action, but not as a start
    or stop, for example.

    The solution is:
    * check if $__OCF_ACTION = monitor and do AMQP checks only then.
    * align sleep (3 sec) for action start loop amongst all of the OCF scripts
    * replace netstat to lsof tool due to /proc FS based approach
      for sockets parsing fits better for high loaded systems with
      many connections. Also, 'lsof -p $pid' will not search for all
      of the existing sockets - which could be tens of housands -
      but only for ones from /proc/$pid/ instead, that is very fast and
      provides no additional load for host OS.

    Closes-bug: #1385755

    Change-Id: I4eba9fd5e3eaa6448590808ea85bdf152b2ef176
    Signed-off-by: Bogdan Dobrelya <email address hidden>