Comment 4 for bug 1827461

Hmm, tried the following:

1) bootstrap a clean 2.5.4 controller;
2) deploy an openstack model;
3) upgrade to 2.6-rc1;
3) scale the model;
4) destroy the model (successful).

I still see a lot of similar errors in the controller log after destroy-model which eventually stop:

2019-05-03 10:43:02 TRACE juju.state.txn database.go:359 ran transaction in 0.022s (retries: 0) []txn.Op{

# ...

                    {
                        Name: "model-uuid",
                        Value: "2108fac7-ea12-402c-84e3-cd7e1e15a762",
                    },
                    {
                        Name: "status",
                        Value: "destroying",
                    },

# ...

2019-05-03 10:44:59 ERROR juju.apiserver.common resource.go:118 error stopping *apiserver.pingTimeout resource: ping timeout
2019-05-03 10:44:59 TRACE juju.state.watcher hubwatcher.go:457 0xc000e1a0f0 got request: watcher.reqUnwatch{key:watcher.watchKey{c:"migrations.status", id:interface {
}(nil)}, ch:(chan<- watcher.Change)(0xc00ddccd20)}
2019-05-03 10:44:59 TRACE juju.state.watcher hubwatcher.go:373 loop finished
2019-05-03 10:44:59 DEBUG juju.state open.go:219 closed state without error