Description of problem:

In 4.1 (ceph-ansible-4.0.23-1.el8cp.noarch) we added a check to fail when either dashboard_admin_password or grafana_admin_password was not explicitly set.

In 4.0 (ceph-ansible-4.0.14-1.el8cp.noarch) these defaults were set:

#dashboard_admin_password: p@ssw0rd
#grafana_admin_password: admin

When upgrading from RHCS 4.0 where the above defaults were used the check causes a failure to fully upgrade (see Bug 1844852).

If you uncomment those settings and use the defaults everything proceeds fine.
If you set dashboard_admin_password to a new password and just uncomment grafana_admin_password and use the default for grafana, everything proceeds fine.

If you set grafana_admin_password to a new password, the following failure is seen:

- When I set "grafana_admin_password: redhat"

```
fatal: [ceph4osd1 -> ceph4osd1]: FAILED! => changed=false 
  cmd:
  - docker
  - exec
  - ceph-mon-ceph4osd1
  - ceph
  - --cluster
  - ceph
  - dashboard
  - grafana
  - dashboards
  - update
  delta: '0:00:01.177480'
  end: '2020-06-07 11:46:52.813923'
  msg: non-zero return code
  rc: 22
  start: '2020-06-07 11:46:51.636443'
  stderr: |-
    Error EINVAL: Traceback (most recent call last):
      File "/usr/share/ceph/mgr/mgr_module.py", line 974, in _handle_command
        return self.handle_command(inbuf, cmd)
      File "/usr/share/ceph/mgr/dashboard/module.py", line 386, in handle_command
        push_local_dashboards()
      File "/usr/share/ceph/mgr/dashboard/grafana.py", line 133, in push_local_dashboards
        retry()
      File "/usr/share/ceph/mgr/dashboard/grafana.py", line 90, in __call__
        result = self.func(*self.args, **self.kwargs)
      File "/usr/share/ceph/mgr/dashboard/grafana.py", line 128, in push
        grafana.push_dashboard(body)
      File "/usr/share/ceph/mgr/dashboard/grafana.py", line 55, in push_dashboard
        response.raise_for_status()
      File "/lib/python3.6/site-packages/requests/models.py", line 940, in raise_for_status
        raise HTTPError(http_error_msg, response=self)
    requests.exceptions.HTTPError: 401 Client Error: Unauthorized for url: http://7.7.7.101:3000/api/dashboards/db

- when I set "grafana_admin_password: admin" it is fine.

Additionally, after upgrade if you try to change grafan_admin_password, it fails with the same issue as above.

See the following original reports:


This is an outgrowth of
https://bugzilla.redhat.com/show_bug.cgi?id=1844852#c2
https://bugzilla.redhat.com/show_bug.cgi?id=1844852#c8


Version-Release number of selected component (if applicable):

RHCS 4.0 cluster upgrade to RHCS 4.1

ceph-ansible-4.0.14-1.el7cp.noarch
ceph-ansible-4.0.23-1.el8cp.noarch


How reproducible: Only tested on Mustafa's test cluster


Steps to Reproduce:
1. see above
2.
3.

Actual results:


Expected results:


Additional info: