Defaulting device names fails to update the database

[Nikola Đipanov]

_default_block_device_names method of the compute manager, would call the conductor block_device_mapping_update method with the wrong arguments, causing a TypeError and ultimately the instance to fail.

This bug happens only when using a driver that does not provid it's own implementation of default_device_names_for_instance, (currently only the libvirt driver does this).

Also affects havana since https://review.openstack.org/#/c/40229/

[Matt Riedemann (mriedem)]

Patch is here: https://review.openstack.org/#/c/62043/

[OpenStack Infra (hudson-openstack)]

Reviewed: https://review.openstack.org/62043
Committed: https://git.openstack.org/cgit/openstack/nova/commit/?id=f6988b3f1ebf56364696ff25448a8018866fd20b
Submitter: Jenkins
Branch: master

commit f6988b3f1ebf56364696ff25448a8018866fd20b
Author: Nikola Dipanov <email address hidden>
Date: Fri Dec 13 17:59:10 2013 +0100

    Fix updating device names when defaulting

    If the virt driver does not support defaulting device names, the default
    behaviour of defaulting them in the manager was calling the conductor
    block_device_mapping_update method with the wrong arguments, which
    caused a TypeError and ultimately a failure to boot the instance.

    This patch fixed the signature and adds a test that exercises it so
    that it does not regress.

    Closes-bug: #1260806
    Change-Id: I6116540c9aca4847afacf80b3d1e0c027c364159