Get stuck in detaching state due redundant bdm

[zhaoleilc]

Description
===========
There will be redundant bdms in nova after a volume
fails to be detached from the server, ie, the libvirt
has succeeded in detaching the disk and attachment
in cinder has disappeared, but bdm in nova is present.
In such scenario, nova compute service will report an
error that device vdb not found in instance if the
volume is detached from the server again.

Expected result
===============
Even though redundant bdms are present, they can be
removed by another volume remove operation. Or these
bdms reside in nova database forever until the database
is directly modified.

[Uggla (rene-ribaud)]

Thanks for submitting this bug, but it is incomplete and can not be treated without the following information. Please add them to the ticket.

- steps to reproduce
- the version of Nova and the novaclient (or os-client)
- logs (on debug level)
- environment details depending on the bug
        libvirt/kvm versions, VMWare version, ...
        storage type (ceph, LVM, GPFS, ...)
        network type (nova-network or neutron)

[Launchpad Janitor (janitor)]

[Expired for OpenStack Compute (nova) because there has been no activity for 60 days.]