Bug #24197
closed"timedatectl', 'set-timezone', 'Etc/UTC'" fails
0%
Description
i guess we set the timezone to "Etc/UTC" at https://github.com/ceph/ceph-cm-ansible/blob/master/roles/common/defaults/main.yml#L2 .
after googling around, it could be a bug in ansible: https://github.com/ansible/ansible/pull/28229 . but i am not sure.
Updated by David Galloway almost 6 years ago
- Assignee set to David Galloway
I've known about this issue for a while but haven't been able to reproduce it to debug it. All I've been able to gather is the login service fails to start sometimes which results in timedatectl failing. This is exclusive to the mira.
Updated by Sebastian Wagner about 5 years ago
Yep. I'm getting
{description: 'rados:mgr/{clusters/{2-node-mgr.yaml openstack.yaml} debug/mgr.yaml
objectstore/bluestore-bitmap.yaml supported-random-distro$/{rhel_latest.yaml}
tasks/insights.yaml}', duration: 687.5767729282379, failure_reason: '{''mira077.front.sepia.ceph.com'':
{''_ansible_parsed'': True, ''stderr_lines'': ["Failed to set time zone: Failed
to activate service ''org.freedesktop.timedate1'': timed out"], ''cmd'': [''timedatectl'',
''set-timezone'', ''Etc/UTC''], ''end'': ''2019-01-30 11:52:34.069291'', ''_ansible_no_log'':
False, ''stdout'': '''', ''changed'': True, ''invocation'': {''module_args'':
{''creates'': None, ''executable'': None, ''chdir'': None, ''_raw_params'': ''timedatectl
set-timezone Etc/UTC'', ''removes'': None, ''warn'': True, ''_uses_shell'': False,
''stdin'': None}}, ''start'': ''2019-01-30 11:52:09.057301'', ''delta'': ''0:00:25.011990'',
''stderr'': "Failed to set time zone: Failed to activate service ''org.freedesktop.timedate1'':
timed out", ''rc'': 1, ''msg'': ''non-zero return code'', ''stdout_lines'': []}}',
owner: scheduled_swagner@teuthology, status: fail, success: false}
Updated by David Galloway about 5 years ago
- Status changed from New to Fix Under Review
So, I don't have a solution for this but I've found a workaround. If the system is rebooted, whatever is broken gets fixed and the playbook can proceed.
What's happening here is something systemd-related. I've found BZs/StackOverflows indicating a user could be missing from /etc/passwd or that a package needs to be reinstalled but our FOG image is static. This happens on some mira and not on others. Therefore, /etc/passwd is identical. The OS image isn't the problem.
https://access.redhat.com/solutions/1543343
https://bugzilla.redhat.com/show_bug.cgi?id=1418886
The solution(s) in the BZ are not the issue we're seeing.
Updated by David Galloway about 5 years ago
- Status changed from Fix Under Review to Resolved
Happy to see this resolved. https://github.com/ceph/ceph-cm-ansible/pull/435