Instances stuck in BUILD should time out by default

[Greg Lucas]

When an instance fails to go ACTIVE after some amount of time, the default behavior should be to put the instance in an ERROR state.

This behavior exists but the default is to disable the time-out such that instances can remain stuck in BUILD forever. For new users this is confusing as there is not a straightforward way to clean out such instances.

This behavior is controlled by these options:

usage_timeout (default=400)
update_status_on_fail (default=False)

[OpenStack Infra (hudson-openstack)]

Fix proposed to branch: master
Review: https://review.openstack.org/105825

[OpenStack Infra (hudson-openstack)]

Reviewed: https://review.openstack.org/105825
Committed: https://git.openstack.org/cgit/openstack/trove/commit/?id=ba21c3a3ab2f139a728625f1f40fe995377addda
Submitter: Jenkins
Branch: master

commit ba21c3a3ab2f139a728625f1f40fe995377addda
Author: Greg Lucas <email address hidden>
Date: Wed Jul 9 14:01:18 2014 -0400

    Change default for update_status_on_fail

    Default 'update_status_on_fail' to True so that instances do not get stuck
    in the BUILD state.

    When this option is set, instances that remain in BUILD for longer than
    the datastore's configured 'usage_timeout' will be considered failed and
    moved to the ERROR state.

    Closes-Bug: 1339826
    Change-Id: I5f4b734830a21a1fbb6f169edb5424bac6f883cd