OSA Queens: libvirtError: Failed to connect socket to '/var/run/libvirt/libvirt-sock': No such file or directory
Affects | Status | Importance | Assigned to | Milestone | |
---|---|---|---|---|---|
OpenStack Compute (nova) |
Invalid
|
Undecided
|
Unassigned | ||
OpenStack-Ansible |
Fix Released
|
High
|
Jesse Pretorius |
Bug Description
I have installed stable/queens 17.0.5 using openstack-ansible on Ubuntu with ODL.
When i reboot compute node, instances go in shut off state.
Log in /var/log/
2018-07-04 10:00:58.496 5519 ERROR nova.virt.
If power fluctuates compute will go down and i have to restart nova-compute on compute node and start all instances.
Is there a fix for this issue?
Status of services after reboot:
root@home:~# service libvirtd status
● libvirtd.service - Virtualization daemon
Loaded: loaded (/lib/systemd/
Active: active (running) since Wed 2018-07-04 10:02:56 IST; 25s ago
Docs: man:libvirtd(8)
https:/
Main PID: 5809 (libvirtd)
Tasks: 17 (limit: 32768)
Memory: 25.0M
CPU: 385ms
CGroup: /system.
└─5809 /usr/sbin/libvirtd
Jul 04 10:02:54 home systemd[1]: Starting Virtualization daemon...
Jul 04 10:02:56 home systemd[1]: Started Virtualization daemon.
root@home:~# service nova-compute status
● nova-compute.
Loaded: loaded (/etc/systemd/
Active: inactive (dead) since Wed 2018-07-04 10:00:58 IST; 2min 30s ago
Process: 5519 ExecStart=
Main PID: 5519 (code=exited, status=0/SUCCESS)
Jul 04 10:00:53 home systemd[1]: Started nova openstack service.
Jul 04 10:00:56 home nova-compute[5519]: /openstack/
Jul 04 10:00:56 home nova-compute[5519]: exception.
summary: |
- libvirtError: Failed to connect socket to '/var/run/libvirt/libvirt- - sock': No such file or directory + OSA Queens: libvirtError: Failed to connect socket to + '/var/run/libvirt/libvirt-sock': No such file or directory |
Changed in openstack-ansible: | |
assignee: | nobody → Kevin Carter (kevin-carter) |
importance: | Undecided → High |
status: | New → Confirmed |
Found the solution for this issue:
libvirtd was starting after nova-compute during boot, so we had this error in nova-compute.
In compute node, for nova-compute service to start first, do the following on ubuntu 16.04
Add "After= libvirtd. service" in /etc/systemd/ system/ nova-compute. service
For instances to resume to previous state after compute reboot do the following: guests_ state_on_ host_boot to True in /etc/nova/nova.conf
set resume_
This should be automated by ansible.
Thanks