volume is in "in use" state when a vmware instance boot from volume is failed

[zhaodan7597]

when creating a vmware instance from a volume is failed, vm goes to error state, the volume is in "in use" state, and after deleting the vm, the state of the volume is still in "in use" so it can't been deleted.

[OpenStack Infra (hudson-openstack)]

Change abandoned by zhaodan7597 (<email address hidden>) on branch: master
Review: https://review.openstack.org/571112

[Matt Riedemann (mriedem)]

Do you have a stacktrace of the failure? Is the not found error happening when deleting the server and the ComputeManager is trying to detach the volume here?

https://github.com/openstack/nova/blob/6bf11e1dc14afad78b11d980c2544a3dc41579ff/nova/compute/manager.py#L2563

[zhaodan7597 (zhaodan7597)]

The not found error happening after the server building failed, not in deleteing.
After building failed, it will rollback resources, and destroy this instance. In nova/virt/vmwareapi/driver.py , code goes to destroy() -->_detach_instance_volumes()-->self._vmops.power_off(), then the power_off() will raise a not found exception.

[OpenStack Infra (hudson-openstack)]

Fix proposed to branch: master
Review: https://review.openstack.org/626095

[OpenStack Infra (hudson-openstack)]

Change abandoned by zhaodan7597 (<email address hidden>) on branch: master
Review: https://review.openstack.org/626095

[OpenStack Infra (hudson-openstack)]

Change abandoned by Matt Riedemann (<email address hidden>) on branch: master
Review: https://review.opendev.org/571112

[Matt Riedemann (mriedem)]

I abandoned the patch, there was a -1 from a core on it for a long time. It looks like addressing that comment would be trivial though if someone wants the patch restored so they can update it.