Fix the kerberos set-up by loosening ubuntu check (#521)
* Fix the kerberos set-up by loosening ubuntu check The cs:ubuntu charm's workload-status-message has changed recently (as the charm was re-written). This PR loosens the application states check to ignore the workload-status-message and just wait for the charm to be active workload-status. This is almost certainly enough, as it's a very simple charm. * Fix tox.ini to pass tests (maybe)
This commit is contained in:
@@ -30,6 +30,26 @@ commands = nosetests --with-coverage --cover-package=zaza.openstack {posargs} {t
|
||||
basepython = python3
|
||||
deps = -r{toxinidir}/requirements.txt
|
||||
|
||||
[testenv:py3.5]
|
||||
basepython = python3.5
|
||||
deps = -r{toxinidir}/requirements.txt
|
||||
|
||||
[testenv:py3.6]
|
||||
basepython = python3.6
|
||||
deps = -r{toxinidir}/requirements.txt
|
||||
|
||||
[testenv:py3.7]
|
||||
basepython = python3.7
|
||||
deps = -r{toxinidir}/requirements.txt
|
||||
|
||||
[testenv:py3.8]
|
||||
basepython = python3.8
|
||||
deps = -r{toxinidir}/requirements.txt
|
||||
|
||||
[testenv:py3.9]
|
||||
basepython = python3.9
|
||||
deps = -r{toxinidir}/requirements.txt
|
||||
|
||||
[testenv:pep8]
|
||||
basepython = python3
|
||||
deps = -r{toxinidir}/requirements.txt
|
||||
|
||||
@@ -113,7 +113,15 @@ def retrieve_and_attach_keytab():
|
||||
'keystone_keytab',
|
||||
tmp_file)
|
||||
|
||||
zaza.model.wait_for_application_states()
|
||||
# cs:ubuntu charm has changed behaviour and we can't rely on the workload
|
||||
# staus message. Thus, ignore it.
|
||||
states = {
|
||||
"ubuntu-test-host": {
|
||||
"workload-status": "active",
|
||||
"workload-status-message": "",
|
||||
}
|
||||
}
|
||||
zaza.model.wait_for_application_states(states=states)
|
||||
zaza.model.block_until_all_units_idle()
|
||||
|
||||
|
||||
|
||||
Reference in New Issue
Block a user