On Wed, Jan 29, 2020 at 10:28 AM Marcin Sobczyk <msobczyk@redhat.com> wrote:

On 1/28/20 5:11 PM, Michal Skrivanek wrote:

On 28 Jan 2020, at 16:14, Martin Perina <mperina@redhat.com> wrote:

On Tue, Jan 28, 2020 at 9:58 AM Yedidyah Bar David <didi@redhat.com> wrote:

Hi all,

Tried yesterday to deploy hosted-engine on current master snapshot, it failed during host-deploy. host-deploy log ends with:

2020-01-27 15:25:11 IST - TASK [ovirt-host-deploy-vdsm : Reconfigure vdsm tool] **************************

/var/log/messages on the host has:

Jan 27 15:25:24 didi-centos8-host python3[12352]: detected unhandled Python exception in '/usr/bin/vdsm-tool'

/var/log/httpd/ansible_runner_error_log on the engine has (I have to get used to searching there, grepped all of /var/log to find it):

[Mon Jan 27 15:25:25.692397 2020] [wsgi:error] [pid 30029] cb_event_handler event_data={u'event': u'runner_on_failed', u'uuid': u'874e5fcc-12e8-419c-b010-d958adfc4aec', 'stdout': u'fatal: [d
idi-centos8-host.lab.eng.tlv2.redhat.com]: FAILED! => {"changed": true, "cmd": "vdsm-tool configure --force", "delta": "0:00:15.833532", "end": "2020-01-27 15:25:25.296212", "msg": "non-zero
return code", "rc": 1, "start": "2020-01-27 15:25:09.462680", "stderr": "Error: Traceback (most recent call last):\\\\n File \\\\"/usr/bin/vdsm-tool\\\\", line 209, in main\\\\n return
tool_command[cmd][\\\\"command\\\\"](*args)\\\\n File \\\\"/usr/lib/python3.6/site-packages/vdsm/tool/__init__.py\\\\", line 40, in wrapper\\\\n func(*args, **kwargs)\\\\n File \\\\"/u
sr/lib/python3.6/site-packages/vdsm/tool/configurator.py\\\\", line 149, in configure\\\\n service.service_start(s)\\\\n File \\\\"/usr/lib/python3.6/site-packages/vdsm/tool/service.py\\
\\", line 193, in service_start\\\\n return _runAlts(_srvStartAlts, srvName)\\\\n File \\\\"/usr/lib/python3.6/site-packages/vdsm/tool/service.py\\\\", line 172, in _runAlts\\\\n \\\\
"%s failed\\\\" % alt.__name__, out, err)\\\\nvdsm.tool.service.ServiceOperationError: <exception str() failed>\\\\n\\\\nDuring handling of the above exception, another exception occurred:\\
\\n\\\\nTraceback (most recent call last):\\\\n File \\\\"/usr/bin/vdsm-tool\\\\", line 224, in <module>\\\\n sys.exit(main())\\\\n File \\\\"/usr/bin/vdsm-tool\\\\", line 214, in main\
\\\n print(\\'Error: \\', e, \\'\\\\\\\\n\\', file=sys.stderr)\\\\n File \\\\"/usr/lib/python3.6/site-packages/vdsm/tool/service.py\\\\", line 75, in __str__\\\\n return \\'\\\\\\\\n\
\'.join(s)\\\\nTypeError: sequence item 1: expected str instance, bytes found", "stderr_lines": ["Error: Traceback (most recent call last):", " File \\\\"/usr/bin/vdsm-tool\\\\", line 209,
in main", " return tool_command[cmd][\\\\"command\\\\"](*args)", " File \\\\"/usr/lib/python3.6/site-packages/vdsm/tool/__init__.py\\\\", line 40, in wrapper", " func(*args, **kwargs
)", " File \\\\"/usr/lib/python3.6/site-packages/vdsm/tool/configurator.py\\\\", line 149, in configure", " service.service_start(s)", " File \\\\"/usr/lib/python3.6/site-packages/vdsm/
tool/service.py\\\\", line 193, in service_start", " return _runAlts(_srvStartAlts, srvName)", " File \\\\"/usr/lib/python3.6/site-packages/vdsm/tool/service.py\\\\", line 172, in _runAl
ts", " \\\\"%s failed\\\\" % alt.__name__, out, err)", "vdsm.tool.service.ServiceOperationError: <exception str() failed>", "", "During handling of the above exception, another exception
occurred:", "", "Traceback (most recent call last):", " File \\\\"/usr/bin/vdsm-tool\\\\", line 224, in <module>", " sys.exit(main())", " File \\\\"/usr/bin/vdsm-tool\\\\", line 214, in
main", " print(\\'Error: \\', e, \\'\\\\\\\\n\\', file=sys.stderr)", " File \\\\"/usr/lib/python3.6/site-packages/vdsm/tool/service.py\\\\", line 75, in __str__", " return \\'\\\\\\\
\n\\'.join(s)", "TypeError: sequence item 1: expected str instance, bytes found"], "stdout": "\\\\nChecking configuration status...\\\\n\\\\nCurrent revision of multipath.conf detected, pres
erving\\\\nabrt is not configured for vdsm\\\\nlibvirtd socket units status: [{\\'Names\\': \\'libvirtd-tls.socket\\', \\'LoadState\\': \\'loaded\\'}, {\\'Names\\': \\'libvirtd-tcp.socket\\'
, \\'LoadState\\': \\'masked\\'}, {\\'Names\\': \\'libvirtd-ro.socket\\', \\'LoadState\\': \\'masked\\'}, {\\'Names\\': \\'libvirtd-admin.socket\\', \\'LoadState\\': \\'masked\\'}, {\\'Names
\\': \\'libvirtd.socket\\', \\'LoadState\\': \\'masked\\'}]\\\\nlibvirtd doesn\\'t use systemd socket activation

This is untrue now, *if* you are sure you’re using advanced virt repo and installed libvirt/qemu from there. Can you confirm? Also, the detection/handling is only fixed in vdsm since last Thu, not in any build (only the tag from today has it, or nightly)
Can you confirm?

+1 with Michal here - this should never happen in real life scenarios.
The most important thing to notice in the log is this part:
libvirtd doesn\\'t use systemd socket activation - one or more of its socket units have been masked
As you rightly noticed in the other devel thread ("Task Start and enable services failed to execute"),
libvirt's socket units are masked by libvirt itself in an upgrade scenario:

- you have old libvirt and vdsm installed
- vdsm configures libvirt to pass the "--listen" flag to the daemon
- you upgrade libvirt and vdsm to a version that uses socket activation
- libvirt installation scripts notice the "--listen" flag and respect it by disabling
its own socket activation mechanism, which is done by masking sockets
- new vdsm version is lost - it sees that there's a new libvirt version that
uses socket activation (because the socket units are there), but they've
been masked in the meantime

We don't support upgrade path by design.

Not sure what you mean here exactly. Upgrade from what to what? Is this documented somewhere? Enforced somehow?

Current vdsm versions require libvirt >= 5.6 - versions that have socket activation [1].

There are indeed some leftovers in 'vdsm-tool' that were trying to handle both scenarios,
but again - no upgrade paths. 'vdsm-tool' should work for fresh installations without any issues.

OK.

I was just wondering, see the other thread, "Why is vdsm enabled by default?".

I'd clarify my current behavior as a developer. I have a VM I use for hosted-engine testing. I take snapshots of it after installing or upgrading packages, but before '--deploy'. Then I try deploy, play with stuff etc., and eventually revert to the latest snapshot, 'dnf update', and take another snapshot, before trying deploy again.

My problem is that on the boot after reversion to a snapshot, vdsm starts and configures stuff, before I try deploy. As you explained, this is currently broken.

I now did above again, but reverted to a snapshot from before installing vdsm. Indeed, now I pass the point that failed above.

I guess a workaround for my own flow above is to disable vdsmd before taking the snapshot...

Thanks, anyway, and sorry for the noise!

Best regards,

[1] https://github.com/oVirt/vdsm/blob/c2735bec16586e1ad2d0ac0242839bfa0e35b4ec/vdsm.spec.in#L225

- one or more of its socket units have been masked\\\\nlibvirt is not configured for vdsm yet\\\\nFAILED: conflicting vdsm and libvirt-qemu tls configuration.\\\\nvdsm.conf with ssl=True requires the following changes:\\\\nlibvirtd.conf: listen_tcp=0, auth_tcp=\\\\"sasl\\\\", listen_tls=1\\\\nqemu.conf: spice_tls=1.\\\\nlvm is configured for vdsm\\\\nManaged volume database is already configured\\\\n\\\\nRunning configure...\\\\nReconfiguration of passwd is done.\\\\nReconfiguration of abrt is done.\\\\nlibvirtd socket units status: [{\\'Names\\': \\'libvirtd-tls.socket\\', \\'LoadState\\': \\'loaded\\'}, {\\'Names\\': \\'libvirtd-tcp.socket\\', \\'LoadState\\': \\'masked\\'}, {\\'Names\\': \\'libvirtd-ro.socket\\', \\'LoadState\\': \\'masked\\'}, {\\'Names\\': \\'libvirtd-admin.socket\\', \\'LoadState\\': \\'masked\\'}, {\\'Names\\': \\'libvirtd.socket\\', \\'LoadState\\': \\'masked\\'}]\\\\nlibvirtd doesn\\'t use systemd socket activation - one or more of its socket units have been masked\\\\nlibvirtd socket units status: [{\\'Names\\': \\'libvirtd-tls.socket\\', \\'LoadState\\': \\'loaded\\'}, {\\'Names\\': \\'libvirtd-tcp.socket\\', \\'LoadState\\': \\'masked\\'}, {\\'Names\\': \\'libvirtd-ro.socket\\', \\'LoadState\\': \\'masked\\'}, {\\'Names\\': \\'libvirtd-admin.socket\\', \\'LoadState\\': \\'masked\\'}, {\\'Names\\': \\'libvirtd.socket\\', \\'LoadState\\': \\'masked\\'}]\\\\nlibvirtd doesn\\'t use systemd socket activation - one or more of its socket units have been masked\\\\nReconfiguration of libvirt is done.", "stdout_lines": ["", "Checking configuration status...", "", "Current revision of multipath.conf detected, preserving", "abrt is not configured for vdsm", "libvirtd socket units status: [{\\'Names\\': \\'libvirtd-tls.socket\\', \\'LoadState\\': \\'loaded\\'}, {\\'Names\\': \\'libvirtd-tcp.socket\\', \\'LoadState\\': \\'masked\\'}, {\\'Names\\': \\'libvirtd-ro.socket\\', \\'LoadState\\': \\'masked\\'}, {\\'Names\\': \\'libvirtd-admin.socket\\', \\'LoadState\\': \\'masked\\'}, {\\'Names\\': \\'libvirtd.socket\\', \\'LoadState\\': \\'masked\\'}]", "libvirtd doesn\\'t use systemd socket activation - one or more of its socket units have been masked", "libvirt is not configured for vdsm yet", "FAILED: conflicting vdsm and libvirt-qemu tls configuration.", "vdsm.conf with ssl=True requires the following changes:", "libvirtd.conf: listen_tcp=0, auth_tcp=\\\\"sasl\\\\", listen_tls=1", "qemu.conf: spice_tls=1.", "lvm is configured for vdsm", "Managed volume database is already configured", "", "Running configure...", "Reconfiguration of passwd is done.", "Reconfiguration of abrt is done.", "libvirtd socket units status: [{\\'Names\\': \\'libvirtd-tls.socket\\', \\'LoadState\\': \\'loaded\\'}, {\\'Names\\': \\'libvirtd-tcp.socket\\', \\'LoadState\\': \\'masked\\'}, {\\'Names\\': \\'libvirtd-ro.socket\\', \\'LoadState\\': \\'masked\\'}, {\\'Names\\': \\'libvirtd-admin.socket\\', \\'LoadState\\': \\'masked\\'}, {\\'Names\\': \\'libvirtd.socket\\', \\'LoadState\\': \\'masked\\'}]", "libvirtd doesn\\'t use systemd socket activation - one or more of its socket units have been masked", "libvirtd socket units status: [{\\'Names\\': \\'libvirtd-tls.socket\\', \\'LoadState\\': \\'loaded\\'}, {\\'Names\\': \\'libvirtd-tcp.socket\\', \\'LoadState\\': \\'masked\\'}, {\\'Names\\': \\'libvirtd-ro.socket\\', \\'LoadState\\': \\'masked\\'}, {\\'Names\\': \\'libvirtd-admin.socket\\', \\'LoadState\\': \\'masked\\'}, {\\'Names\\': \\'libvirtd.socket\\', \\'LoadState\\': \\'masked\\'}]", "libvirtd doesn\\'t use systemd socket activation - one or more of its socket units have been masked", "Reconfiguration of libvirt is done."]}', 'counter': 40, u'pid': 31685, u'created': u'2020-01-27T13:25:25.683262', 'end_line': 35, 'runner_ident': '5b95ec42-4108-11ea-8afe-001a4a16027a', 'start_line': 34, u'event_data': {u'play_pattern': u'all', u'play': u'all', u'event_loop': None, u'task_args': u'', u'remote_addr': u'didi-centos8-host.lab.eng.tlv2.redhat.com', u'res': {u'stderr_lines': [u'Error: Traceback (most recent call last):', u' File "/usr/bin/vdsm-tool", line 209, in main', u' return tool_command[cmd]["command"](*args)', u' File "/usr/lib/python3.6/site-packages/vdsm/tool/__init__.py", line 40, in wrapper', u' func(*args, **kwargs)', u' File "/usr/lib/python3.6/site-packages/vdsm/tool/configurator.py", line 149, in configure', u' service.service_start(s)', u' File "/usr/lib/python3.6/site-

The exception seems to be cut somehow, but anyway Marcin, could you please take a look?

[Mon Jan 27 15:25:25.693655 2020] [wsgi:error] [pid 30029] cb_event_handler event_data={u'event': u'playbook_on_stats', u'uuid': u'2c8d843c-c1ad-41c0-940e-88354387b725', 'stdout': u'\\r\\nPLAY RECAP *********************************************************************\\r\\ndidi-centos8-host.lab.eng.tlv2.redhat.com : ok=11 changed=0 unreachable=0 failed=1 skipped=1 rescued=0 ignored=0 \\r\\n', 'counter': 41, u'pid': 31685, u'created': u'2020-01-27T13:25:25.688541', 'end_line': 39, 'runner_ident': '5b95ec42-4108-11ea-8afe-001a4a16027a', 'start_line': 35, u'event_data': {u'ignored': {}, u'skipped': {u'didi-centos8-host.lab.eng.tlv2.redhat.com': 1}, u'ok': {u'didi-centos8-host.lab.eng.tlv2.redhat.com': 11}, u'artifact_data': {}, u'rescued': {}, u'changed': {}, u'pid': 31685, u'dark': {}, u'playbook_uuid': u'ff508986-9930-47a2-b17b-61a2f2dae905', u'playbook': u'ovirt-host-deploy.yml', u'failures': {u'didi-centos8-host.lab.eng.tlv2.redhat.com': 1}, u'processed': {u'didi-centos8-host.lab.eng.tlv2.redhat.com': 1}}, u'parent_uuid': u'ff508986-9930-47a2-b17b-61a2f2dae905'}

If I now manually run on the host 'vdsm-tool configure --force', it does not fail.

I am now going to try deploy again after reverting to a snapshot I took before deploy.

Best regards,

--

Didi

_______________________________________________
Devel mailing list -- devel@ovirt.org
To unsubscribe send an email to devel-leave@ovirt.org
Privacy Statement: https://www.ovirt.org/site/privacy-policy/
oVirt Code of Conduct: https://www.ovirt.org/community/about/community-guidelines/
List Archives: https://lists.ovirt.org/archives/list/devel@ovirt.org/message/Q2VEW5NXQPVGPUAA63QGXG7ZTE4MZWA5/

--

Martin Perina
Manager, Software Engineering
Red Hat Czech s.r.o.

_______________________________________________
Devel mailing list -- devel@ovirt.org
To unsubscribe send an email to devel-leave@ovirt.org
Privacy Statement: https://www.ovirt.org/site/privacy-policy/
oVirt Code of Conduct: https://www.ovirt.org/community/about/community-guidelines/
List Archives: https://lists.ovirt.org/archives/list/devel@ovirt.org/message/A44YHUFU54HAOPNL6NLRVU6XJQGIVPON/

Didi