VMware: _get_power_state does not set correct power state

[Gary Kotton]

In the event that an instance is deleted during the driver get_info method then the instance may not be correctly set. This is due to the fact that the VMware driver does not raise the correct exception.

[OpenStack Infra (hudson-openstack)]

Fix proposed to branch: master
Review: https://review.openstack.org/191558

[OpenStack Infra (hudson-openstack)]

Reviewed: https://review.openstack.org/191558
Committed: https://git.openstack.org/cgit/openstack/nova/commit/?id=ca6a563715428bfe56e85e1a67e77da99de2939a
Submitter: Jenkins
Branch: master

commit ca6a563715428bfe56e85e1a67e77da99de2939a
Author: Gary Kotton <email address hidden>
Date: Sun Jun 14 02:21:37 2015 -0700

    VMware: ensure that get_info raises the correct exception

    In the event that a instance is deleted during the get_info method
    then the method would not raise the correct exception.

    If the instance no longer exists then the VC will raise
    a 'ManagedObjectNotFound' exception.

    Note that the oslo.vmware patch that supports this exception is:
    https://review.openstack.org/#/c/191553/

    This exception handling is treated by using vexc.register_fault_class
    to support the local exception (until replaced by oslo.vmware)

    Co-Authored-By: Radoslav Gerganov <email address hidden>

    Change-Id: I2d4b131789319cfd1322cbc95edd7822114532e1
    Closes-bug: #1464979