[openstack-ansible][nova] os-nova-install.yml playbook fails after being unable to restart libvirtd.service on compute nodes

Oliver Wenz oliver.wenz at dhbw-mannheim.de
Fri Nov 6 15:43:51 UTC 2020


To enable live migration on an ussuri (deployed using OpenStack Ansible)
I set up a shared filesystem (ocfs2) on the compute hosts. Then I added
the following to /etc/openstack_deploy/user_variables.yml and reran the
os-nova-install.yml playbook:

```
nova_nova_conf_overrides:
  libvirt:
      iscsi_use_multipath: true

nova_system_user_uid: 999
nova_system_group_gid: 999

nova_novncproxy_vncserver_listen: 0.0.0.0


# Disable libvirtd's TLS listener
nova_libvirtd_listen_tls: 0

# Enable libvirtd's plaintext TCP listener
nova_libvirtd_listen_tcp: 1

nova_libvirtd_auth_tcp: none
```

The playbook failed with the following message:

```
RUNNING HANDLER [os_nova : Restart libvirt-bin]
***************************************************************************
        fatal: [compute1]: FAILED! => {"changed": false, "msg": "Unable
to restart service libvirtd: Job for libvirtd.service failed because the
control process exited with error code.\nSee \"systemctl status
libvirtd.service\" and \"journalctl -xe\" for details.\n"}
```

(similar for the other compute nodes)


Running 'systemctl status libvirtd.service' revealed that the service
failed to start (code=exited, status=6).

Here's the 'journalctl -xe' output:


```
Nov 06 12:42:20 bc1bl12 audit[1]: SERVICE_START pid=1 uid=0
auid=4294967295 ses=4294967295 msg='unit=libvirtd comm="systemd"
exe="/lib/systemd/systemd" hostname=? addr=? terminal=? res=failed'
Nov 06 12:42:20 bc1bl12 systemd[1]: libvirtd.service: Service hold-off
time over, scheduling restart.
Nov 06 12:42:20 bc1bl12 systemd[1]: libvirtd.service: Scheduled restart
job, restart counter is at 5.
-- Subject: Automatic restarting of a unit has been scheduled
-- Defined-By: systemd
-- Support: http://www.ubuntu.com/support
-- 
-- Automatic restarting of the unit libvirtd.service has been scheduled,
as the result for
-- the configured Restart= setting for the unit.
Nov 06 12:42:20 bc1bl12 systemd[1]: Stopped Virtualization daemon.
-- Subject: Unit libvirtd.service has finished shutting down
-- Defined-By: systemd
-- Support: http://www.ubuntu.com/support
-- 
-- Unit libvirtd.service has finished shutting down.
Nov 06 12:42:20 bc1bl12 audit[1]: SERVICE_START pid=1 uid=0
auid=4294967295 ses=4294967295 msg='unit=libvirtd comm="systemd"
exe="/lib/systemd/systemd" hostname=? addr=? terminal=? res=success'
Nov 06 12:42:20 bc1bl12 audit[1]: SERVICE_STOP pid=1 uid=0
auid=4294967295 ses=4294967295 msg='unit=libvirtd comm="systemd"
exe="/lib/systemd/systemd" hostname=? addr=? terminal=? res=success'
Nov 06 12:42:20 bc1bl12 systemd[1]: libvirtd.service: Start request
repeated too quickly.
Nov 06 12:42:20 bc1bl12 systemd[1]: libvirtd.service: Failed with result
'exit-code'.
Nov 06 12:42:20 bc1bl12 systemd[1]: Failed to start Virtualization daemon.
-- Subject: Unit libvirtd.service has failed
-- Defined-By: systemd
-- Support: http://www.ubuntu.com/support
-- 
-- Unit libvirtd.service has failed.
-- 
-- The result is RESULT.
Nov 06 12:42:20 bc1bl12 systemd[1]: libvirtd-admin.socket: Failed with
result 'service-start-limit-hit'.
Nov 06 12:42:20 bc1bl12 systemd[1]: libvirtd.socket: Failed with result
'service-start-limit-hit'.
Nov 06 12:42:20 bc1bl12 systemd[1]: libvirtd-ro.socket: Failed with
result 'service-start-limit-hit'.
Nov 06 12:42:20 bc1bl12 systemd[1]: Closed Libvirt local read-only socket.
-- Subject: Unit libvirtd-ro.socket has finished shutting down
-- Defined-By: systemd
-- Support: http://www.ubuntu.com/support
-- 
-- Unit libvirtd-ro.socket has finished shutting down.
```


The /etc/libvirt/libvirtd.conf file written by OpenStack Ansible looks
like this:

```
listen_tls = 0
listen_tcp = 1
unix_sock_group = "libvirt"
unix_sock_ro_perms = "0777"
unix_sock_rw_perms = "0770"
auth_unix_ro = "none"
auth_unix_rw = "none"
auth_tcp = "none"
```

I've encountered mentions of /etc/default/libvirt-bin in documentation
for older OpenStack versions and I'm unsure if something went wrong with
the playbooks, because the file doesn't exist on my compute nodes.

Any ideas that might help are highly appreciated!

Kind regards,
Oliver





More information about the openstack-discuss mailing list