To Reproduce Steps to reproduce the behavior:

1. Deploy an openstack cluster with multiple cinder backends and each with a different 'volume_backend_name'
2. Create an audit with the zone migration strategy to migrate a volume from a pool in one backend to a pool in another backend

Expected behavior

• The volume is migrated to the destination pool. Instead, an error is raised as watcher considers that the volume will not be able to be migrated to the destination pool, as that host can't schedule volumes of the volume type in question.

Bug impact

• Any volume migration in a cluster with the storage setup described will fail.

Known workaround

• Configure volume types with a volume_backend_name, and all cinder backends that the user wants to migrate volumes between with the same volume_backend_name