[mos] TRACE ceilometer.alarm.service HTTPUnauthorized: HTTPUnauthorized (HTTP 401)

[Anastasia Palkina]

"build_id": "2014-06-19_00-31-14",
"mirantis": "yes",
"build_number": "260",
"ostf_sha": "319d3858b24a962e87191a8d5fd4c1e46e1cd959",
"nailgun_sha": "6a995917e010edb247160ee5333b15fa564ed361",
"production": "docker",
"api": "1.0",
"fuelmain_sha": "7089ffea68000b90e5ba85dbfec64e18a6216eea",
"astute_sha": "694b5a55695e01e1c42185bfac9cc7a641a9bd48",
"release": "5.1",
"fuellib_sha": "a6fe0b4f41e6e251e44fb35f771c48f353e33530"

1. Create new environment (Ubuntu, HA mode)
2. Choose GRE segmentation
3. Choose both Ceph
4. Choose Sahara and Ceilometer
5. Add 3 controllers+ceph, 4 computes, 2 ceph, 1 mongo
6. Start deployment. It was successful
7. But there is error in ceilometer-alarm-evaluator.log on first controller (node-1):

2014-06-19 13:41:34 ERROR

ceilometer.alarm.service [-] alarm evaluation cycle failed
2014-06-19 12:41:34.039 18244 TRACE ceilometer.alarm.service Traceback (most recent call last):
2014-06-19 12:41:34.039 18244 TRACE ceilometer.alarm.service File "/usr/lib/python2.7/dist-packages/ceilometer/alarm/service.py", line 91, in _evaluate_assigned_alarms
2014-06-19 12:41:34.039 18244 TRACE ceilometer.alarm.service alarms = self._assigned_alarms()
2014-06-19 12:41:34.039 18244 TRACE ceilometer.alarm.service File "/usr/lib/python2.7/dist-packages/ceilometer/alarm/service.py", line 134, in _assigned_alarms
2014-06-19 12:41:34.039 18244 TRACE ceilometer.alarm.service 'value': True}])
2014-06-19 12:41:34.039 18244 TRACE ceilometer.alarm.service File "/usr/lib/python2.7/dist-packages/ceilometerclient/v2/alarms.py", line 70, in list
2014-06-19 12:41:34.039 18244 TRACE ceilometer.alarm.service return self._list(options.build_url(self._path(), q))
2014-06-19 12:41:34.039 18244 TRACE ceilometer.alarm.service File "/usr/lib/python2.7/dist-packages/ceilometerclient/common/base.py", line 58, in _list
2014-06-19 12:41:34.039 18244 TRACE ceilometer.alarm.service resp, body = self.api.json_request('GET', url)
2014-06-19 12:41:34.039 18244 TRACE ceilometer.alarm.service File "/usr/lib/python2.7/dist-packages/ceilometerclient/common/http.py", line 190, in json_request
2014-06-19 12:41:34.039 18244 TRACE ceilometer.alarm.service resp, body_iter = self._http_request(url, method, **kwargs)
2014-06-19 12:41:34.039 18244 TRACE ceilometer.alarm.service File "/usr/lib/python2.7/dist-packages/ceilometerclient/common/http.py", line 173, in _http_request
2014-06-19 12:41:34.039 18244 TRACE ceilometer.alarm.service raise exc.from_response(resp, ''.join(body_iter))
2014-06-19 12:41:34.039 18244 TRACE ceilometer.alarm.service HTTPUnauthorized: HTTPUnauthorized (HTTP 401)
2014-06-19 12:41:34.039 18244 TRACE ceilometer.alarm.service

[Dina Belova (dbelova)]

Changed to 'medium' as if is was actually working MOS env, I see no reason to mark this as 'high'.

[Dina Belova (dbelova)]

Currently we're working on the 5.0.1, so we can't confirm (or not) this right now, we'll do that when it'll be the opportunity after 5.0.1 work will be finished.

[Dina Belova (dbelova)]

This is non-related to the working process issue - that's because during the OS the installation alarm evaluator starts before keystone begins actually working - that leads to the no auth passed in the https://github.com/openstack/ceilometer/blob/master/ceilometer/alarm/service.py#L130 that's called from the start method as a periodic task. That seems to be fixed with the alarm evaluator started *after* the keystone.

[Dina Belova (dbelova)]

Due to the discussions in the https://bugs.launchpad.net/fuel/+bug/1346335 - logs like that are normal for the not fully started env - if we'll refactor all deployment process not to see them in ceilometer logs, we'll definitely have problems with other services.