LoopingCalls in libvirt return immediately

[Vish Ishaya]

The timers in libvirt waiting for other commands do not actually get started. This means any code that runs after a looping call get executed immediately leading to odd errors like the following on delete:

libvir: QEMU error : Domain not found: no domain with matching name 'instance-00000002'2012-06-19 23:46:32 ERROR nova.virt.libvirt.connection [-] [instance: 2ab3fe58-0b89-4abe-87f5-71c396e43ffe] During wait destroy, instance disappeared.2012-06-19 23:46:33 INFO nova.virt.libvirt.connection [req-54695e07-82fe-483d-94f6-7f13b766e0a3 demo demo] [instance: 2ab3fe58-0b89-4abe-87f5-71c396e43ffe] Deleting instance files /home/vagrant/instance-00000002

[OpenStack Infra (hudson-openstack)]

Fix proposed to branch: master
Review: https://review.openstack.org/8731

[OpenStack Infra (hudson-openstack)]

Reviewed: https://review.openstack.org/8731
Committed: http://github.com/openstack/nova/commit/23ec8c915c900fcafd1683da000817ef4497bf93
Submitter: Jenkins
Branch: master

commit 23ec8c915c900fcafd1683da000817ef4497bf93
Author: Vishvananda Ishaya <email address hidden>
Date: Wed Jun 20 00:44:25 2012 +0000

    Make libvirt LoopingCalls actually wait()

     * Adds wait() to all the LoopingCall timers
     * Fixes loopingCalls in baremetal/proxy.py as well
     * Includes failing test to verify result from destroy
     * Fixes tests depending on improper behavior
     * Fixes bug 1015355

    Change-Id: I9e1914c446170e49f0aab76c7745b55d12132425