Redundant key list generation on dict iteration

[Zhongyue Luo]

There are some cases in the source where redundant key lists are generated for dict iteration.

Using iteritems() lets us iterate a dict more efficiently.

[Zhongyue Luo (zyluo)]

https://review.openstack.org/#/c/129824/

[OpenStack Infra (hudson-openstack)]

Reviewed: https://review.openstack.org/129824
Committed: https://git.openstack.org/cgit/openstack/sahara/commit/?id=eaace6b740b4e112d9cb4c024a2fe6a6765e0f37
Submitter: Jenkins
Branch: master

commit eaace6b740b4e112d9cb4c024a2fe6a6765e0f37
Author: Zhongyue Luo <email address hidden>
Date: Tue Oct 21 14:46:04 2014 +0800

    Fix dict iteration usage

    When iterating a dict, it is redundant to create a list of keys
    using the keys() method. Using iteritems() lets us iterate a dict more
    efficiently.

    Closes-Bug: #1384990

    Change-Id: If8ffcf19129430506020adb9716a7dfdda9d98f5