Hi,
The following happened on an HA-enabled (3 units) 2.3.8 controller, which had applications deployed on the controller units (with juju storage tied to these applications), running in openstack :
$ juju destroy-controller -y <name> --destroy-storage
Destroying controller
Waiting for hosted model resources to be reclaimed
All hosted models reclaimed, cleaning up controller machines
ERROR destroying controller model: destroying storage: destroying volumes: detaching volume 59a3b67e-b30d-4330-a467-8f66542d580e from server 0691094d-dc9d-4d57-82ff-f56d6fd1afc7: failed to delete volume attachment
caused by: request (http://1.2.3.4:8774/v2/afce86f993d943baacfb6509ced93bec/servers/0691094d-dc9d-4d57-82ff-f56d6fd1afc7/os-volume_attachments/59a3b67e-b30d-4330-a467-8f66542d580e) returned unexpected status: 409; error info: {"conflictingRequest": {"message": "Cannot 'detach_volume' while instance is in task_state deleting", "code": 409}}, detaching volume 8b7c8fa8-805b-4c3f-a1a6-38c531e9151c from server 35fafc77-bb88-455d-8adb-46295aaaeeac: failed to delete volume attachment
caused by: request (http://1.2.3.4:8774/v2/afce86f993d943baacfb6509ced93bec/servers/35fafc77-bb88-455d-8adb-46295aaaeeac/os-volume_attachments/8b7c8fa8-805b-4c3f-a1a6-38c531e9151c) returned unexpected status: 409; error info: {"conflictingRequest": {"message": "Cannot 'detach_volume' while instance is in task_state deleting", "code": 409}}
I believe juju should wait for instance deletion to finish before trying to detach/destroy the volumes.
Thanks
The end result is that 3 out of 3 nova instances got deleted, but only 1 out of 3 cinder volumes were deleted.