Code missing from error message when quota-breaching instance launched via OS API

[Eoghan Glynn]

When an instance launch is attempted via the OpenStack API, but the instances quota is already fully utilized, the error message in the 413 overLimit response is not correctly marshalled:

 {u'overLimit': {u'message': u'Quota exceeded: code=%(code)s', u'code': 413, u'retryAfter': 0}}

The 'code=%(code)s' string lacks parameter subsititution so is unclear as to the actual quota that has been breached (e.g. it could be floating IP as opposed to instance stravation).

Whereas when attempting the instance launch via the EC2 API, the correct code is included in the 413 response:

  Quota exceeded: code=InstanceLimitExceeded

[OpenStack Infra (hudson-openstack)]

Fix proposed to branch: master
Review: https://review.openstack.org/5151

[OpenStack Infra (hudson-openstack)]

Reviewed: https://review.openstack.org/5151
Committed: http://github.com/openstack/nova/commit/b5f75253048438891a8740a1ea4719e5f3208ccb
Submitter: Jenkins
Branch: master

commit b5f75253048438891a8740a1ea4719e5f3208ccb
Author: Eoghan Glynn <email address hidden>
Date: Fri Mar 9 15:15:23 2012 +0000

    Sub in InstanceLimitExceeded in overLimit message

    Fixes bug 950768

    Previously the error message in the 413 overLimit response returned
    by the OS compute API was missing a parameter subsitution indicating
    the actual quota that was breeched:

      'Quota exceeded: code=%(code)s'

    This is now properly reported as:

      'Quota exceeded: code=InstanceLimitExceeded'

    Also a potentially useful log message is no longer discarded.

    Change-Id: I7e1f8b078329a6530bb7d7ea3b006ff578cdcdd3