This bug was initially created as a copy of Bug #1774243

I am copying this bug because: 

When faced with a deployment recovery involving the need to delete placement resource providers and recover instance allocations in placement, having access to the 'nova-manage heal_allocations' CLI would be helpful. The CLI was introduced in the Rocky release (OSP14) and we can backport it to OSP13.

Description of problem:
We have a failed Compute node. When we try to migrate VM's from it, we receive the following traceback:

{u'message': u'Failed to create resource provider compute-0', u'code': 500, u'details': u'Traceback (most recent call last):
  File "/usr/lib/python2.7/site-packages/nova/compute/manager.py", line 202, in decorated_function
    return function(self, context, *args, **kwargs)
  File "/usr/lib/python2.7/site-packages/nova/compute/manager.py", line 2920, in rebuild_instance
    migration=migration)
  File "/usr/lib/python2.7/site-packages/oslo_concurrency/lockutils.py", line 274, in inner
    return f(*args, **kwargs)
  File "/usr/lib/python2.7/site-packages/nova/compute/resource_tracker.py", line 246, in rebuild_claim
    limits=limits, image_meta=image_meta)
  File "/usr/lib/python2.7/site-packages/nova/compute/resource_tracker.py", line 355, in _move_claim
    self._update(elevated, cn)
  File "/usr/lib/python2.7/site-packages/nova/compute/resource_tracker.py", line 904, in _update
    inv_data,
  File "/usr/lib/python2.7/site-packages/nova/scheduler/client/__init__.py", line 68, in set_inventory_for_provider
    parent_provider_uuid=parent_provider_uuid,
  File "/usr/lib/python2.7/site-packages/nova/scheduler/client/__init__.py", line 37, in __run_method
    return getattr(self.instance, __name)(*args, **kwargs)
  File "/usr/lib/python2.7/site-packages/nova/scheduler/client/report.py", line 1104, in set_inventory_for_provider
    parent_provider_uuid=parent_provider_uuid)
  File "/usr/lib/python2.7/site-packages/nova/scheduler/client/report.py", line 673, in _ensure_resource_provider
    name=name or uuid)
ResourceProviderCreationFailed: Failed to create resource provider compute-0
', u'created': u'2019-11-12T22:51:43Z'}

Version-Release number of selected component (if applicable):
RHOSP13 z7.
(I'll get the exact nova version and provide it as a comment)

How reproducible:
Every time we try to migrate in this environment

Steps to Reproduce:
1. nova evacuate UUID  (from the failed hypervisor)
2. 
3.

Actual results:

{u'message': u'Failed to create resource provider compute-0', u'code': 500, u'details': u'Traceback (most recent call last):
  File "/usr/lib/python2.7/site-packages/nova/compute/manager.py", line 202, in decorated_function
    return function(self, context, *args, **kwargs)
  File "/usr/lib/python2.7/site-packages/nova/compute/manager.py", line 2920, in rebuild_instance
    migration=migration)
  File "/usr/lib/python2.7/site-packages/oslo_concurrency/lockutils.py", line 274, in inner
    return f(*args, **kwargs)
  File "/usr/lib/python2.7/site-packages/nova/compute/resource_tracker.py", line 246, in rebuild_claim
    limits=limits, image_meta=image_meta)
  File "/usr/lib/python2.7/site-packages/nova/compute/resource_tracker.py", line 355, in _move_claim
    self._update(elevated, cn)
  File "/usr/lib/python2.7/site-packages/nova/compute/resource_tracker.py", line 904, in _update
    inv_data,
  File "/usr/lib/python2.7/site-packages/nova/scheduler/client/__init__.py", line 68, in set_inventory_for_provider
    parent_provider_uuid=parent_provider_uuid,
  File "/usr/lib/python2.7/site-packages/nova/scheduler/client/__init__.py", line 37, in __run_method
    return getattr(self.instance, __name)(*args, **kwargs)
  File "/usr/lib/python2.7/site-packages/nova/scheduler/client/report.py", line 1104, in set_inventory_for_provider
    parent_provider_uuid=parent_provider_uuid)
  File "/usr/lib/python2.7/site-packages/nova/scheduler/client/report.py", line 673, in _ensure_resource_provider
    name=name or uuid)
ResourceProviderCreationFailed: Failed to create resource provider compute-0
', u'created': u'2019-11-12T22:51:43Z'}

Expected results:
It should be able to find the resource provider and migrate the VM.

Additional info:
We fail here:
https://opendev.org/openstack/nova/src/branch/stable/queens/nova/scheduler/client/report.py#L659-L673

So I assume it fails to get / refresh the existing resource provider and moves into the if not statement:

        rps_to_refresh = self._get_providers_in_tree(context, uuid)
        if not rps_to_refresh: