Comment 10 for bug 985489

I find this issue in my stable folsom, but unortunately it occurs randomly, and my libvirt version is 0.9.12-5(debian wheezy).
log in the nova-compute.log:
2013-02-26 02:48:45 DEBUG nova.manager [-] Running periodic task ComputeManager.update_available_resource from (pid=35879) periodic_tasks /usr/local/lib/python2.7/dist-packages/nova/manager.py:200
2013-02-26 02:48:45 DEBUG nova.utils [-] Got semaphore "compute_resources" for method "update_available_resource"... from (pid=35879) inner /usr/local/lib/python2.7/dist-packages/nova/utils.py:803
==============and then the periodic task stopped until============
2013-02-26 09:37:48 ERROR nova.manager [-] Error during ComputeManager.update_available_resource: Domain not found: no domain with matching id 1519
......
2013-02-26 09:37:48 DEBUG nova.manager [-] Running periodic task ComputeManager._report_driver_status from (pid=35879) periodic_tasks /usr/local/lib/python2.7/dist-packages/nova/manager.py:200
2013-02-26 09:37:48 INFO nova.compute.manager [-] Updating host status
2013-02-26 09:37:48 DEBUG nova.virt.libvirt.driver [-] Updating host stats from (pid=35879) update_status /usr/local/lib/python2.7/dist-packages/nova/virt/libvirt/driver.py:3602
==============and then the periodic task stopped again until I restart nova-compute service============

At the second time when this issue occurs, I try to restart libvirtd and don't restart nova-compute, the nova-compute continues to running after I restart libvirtd! so I believe this issue is resulted from libvirtd.
We may need to add a timeout when nova-compute call libvirt apis.