stable/new: resource tracker exception

[Gary Kotton]

2016-10-06 20:20:30.611 8718 WARNING nova.compute.monitors [req-6606069f-7dd1-41ad-997e-3b4b8ca276e4 - -] Excluding nova.compute.monitors.cpu monitor virt_driver. Not in the list of enabled monitors (CONF.compute_monitors).
2016-10-06 20:20:30.612 8718 INFO nova.compute.resource_tracker [req-6606069f-7dd1-41ad-997e-3b4b8ca276e4 - -] Auditing locally available compute resources for node kvm-comp1
2016-10-06 20:20:31.987 8718 ERROR nova.compute.manager [req-6606069f-7dd1-41ad-997e-3b4b8ca276e4 - -] Error updating resources for node kvm-comp1.
2016-10-06 20:20:31.987 8718 ERROR nova.compute.manager Traceback (most recent call last):
2016-10-06 20:20:31.987 8718 ERROR nova.compute.manager File "/opt/stack/nova/nova/compute/manager.py", line 6410, in update_available_resource_for_node
2016-10-06 20:20:31.987 8718 ERROR nova.compute.manager rt.update_available_resource(context)
2016-10-06 20:20:31.987 8718 ERROR nova.compute.manager File "/opt/stack/nova/nova/compute/resource_tracker.py", line 513, in update_available_resource
2016-10-06 20:20:31.987 8718 ERROR nova.compute.manager resources = self.driver.get_available_resource(self.nodename)
2016-10-06 20:20:31.987 8718 ERROR nova.compute.manager File "/opt/stack/nova/nova/virt/libvirt/driver.py", line 5301, in get_available_resource
2016-10-06 20:20:31.987 8718 ERROR nova.compute.manager disk_over_committed = self._get_disk_over_committed_size_total()
2016-10-06 20:20:31.987 8718 ERROR nova.compute.manager File "/opt/stack/nova/nova/virt/libvirt/driver.py", line 6846, in _get_disk_over_committed_size_total
2016-10-06 20:20:31.987 8718 ERROR nova.compute.manager local_instances[guest.uuid], bdms[guest.uuid])
2016-10-06 20:20:31.987 8718 ERROR nova.compute.manager KeyError: 'c7704e45-e60a-4169-a143-4668e47f0b4b'
2016-10-06 20:20:31.987 8718 ERROR nova.compute.manager

[Matthew Booth (mbooth-9)]

The immediate cause here seems to be that there is no entry in bdms for the instance uuid. There are 2 potential sources of KeyError on that line, and we know it's not the other one because it's the guard condition of that code block.

So this line:

        # Get bdms by instance uuids
        bdms = objects.BlockDeviceMappingList.bdms_by_instance_uuid(
            ctx, instance_uuids)

returned no bdms for this instance. I would expect there to be a BDM for at least the root disk. Can anybody confirm?

[OpenStack Infra (hudson-openstack)]

Fix proposed to branch: master
Review: https://review.openstack.org/382833

[Matt Riedemann (mriedem)]

This doesn't seem critical.

[OpenStack Infra (hudson-openstack)]

Change abandoned by Matt Riedemann (<email address hidden>) on branch: master
Review: https://review.openstack.org/382833