Skip to content

Exercise 4. Restarting nova_libvirt without stopping instances

Goal: restart the container that runs libvirt, the way a reconfigure or an upgrade does, and measure that the instances keep running. Time: about 10 minutes. Runbook: I need to restart nova_libvirt on a compute. Its instances keep running

Before you start

Read before every exercise. You need an instance you can ping from the lab VM: create one with a floating IP. The lab's default security group, which the instance gets, already allows ping and SSH.

openstack server create --image cirros --flavor m1.tiny --network demo-net --key-name mykey --wait ex4-test
openstack floating ip create public1 -f value -c floating_ip_address
openstack server add floating ip ex4-test <floating-ip>

The second command prints only the address: that is <floating-ip>. Wait about 30 seconds (the instance is still booting, and the floating IP takes a few seconds to apply), then:

ping -c 3 <floating-ip>

If it still does not answer after a minute, check that the instance is ACTIVE and that the floating IP is attached (openstack server show ex4-test -c status -c addresses).

Note the domain ID libvirt gives the instance:

sudo docker exec nova_libvirt virsh list

1. Do it

In a second SSH session, leave a ping running:

ping -i 0.5 <floating-ip>

Then, in the first session, follow the runbook's Resolution step:

⚠️ State-changing. Restarts libvirt. With Kolla-Ansible's default settings the runbook's lab saw no instance stop; check the runbook's preconditions before doing this on any platform someone relies on.

sudo systemctl restart kolla-nova_libvirt-container.service

2. What you should see

  • The ping keeps answering. On the runbook's lab: 174 requests in the first minute, none lost.
  • sudo docker exec nova_libvirt virsh list shows the same domain ID as before: libvirt took the running instance back; it did not start it again.

Before reading on: where does the instance actually run, if libvirt can restart under it?

3. Read the runbook

Read its Diagnosis and Validity and limits: why the instances survive, and which settings would change that.

4. Check

Follow the runbook's Verification.

5. Clean up

Stop the ping (Ctrl-C), then:

openstack server delete --wait ex4-test
openstack floating ip delete <floating-ip>

Done with the lab for today? Delete everything: a running lab keeps costing.