oVirt Cluster Upgrade

The ovirt.cluster-upgrade role iterates through all the hosts in a cluster and upgrades them.

Note

Please note that when installing this role from Ansible Galaxy you are instructed to run following command:

$ ansible-galaxy install ovirt.cluster-upgrade

This will download the role to the directory with the same name as you specified on the command line, in this case ovirt.cluster-upgrade. But note that it is case sensitive, so if you specify for example OVIRT.cluster-upgrade it will download the same role, but it will add it to the directory named OVIRT.cluster-upgrade, so you later always have to use this role with upper case prefix. So be careful how you specify the name of the role on command line.

For the RPM installation we install three legacy names ovirt.cluster-upgrade, oVirt.cluster-upgrade and ovirt-cluster-upgrade. So you can use any of this name. This documentation and examples in this repository are using name ovirt.cluster-upgrade. oVirt.cluster-upgrade and ovirt-cluster-upgrade role names are deprecated.

Requirements

• Ansible version 2.9 or higher
• Python SDK version 4.3 or higher

Role Variables

Name	Default value
cluster_name	Default	Name of the cluster to be upgraded.
stopped_vms	UNDEF	List of virtual machines to stop before upgrading.
stop_non_migratable_vms alias: stop_pinned_to_host_vms	false	Specify whether to stop virtual machines pinned to the host being upgraded. If true, the pinned non-migratable virtual machines will be stopped and host will be upgraded, otherwise the host will be skipped.
upgrade_timeout	3600	Timeout in seconds to wait for host to be upgraded.
host_statuses	[UP]	List of host statuses. If a host is in any of the specified statuses then it will be upgraded.
host_names	[*]	List of host names to be upgraded.
check_upgrade	false	If true, run check_for_upgrade action on all hosts before executing upgrade on them. If false, run upgrade only for hosts with available upgrades and ignore all other hosts.
reboot_after_upgrade	true	If true reboot hosts after successful upgrade.
use_maintenance_policy	true	If true the cluster policy will be switched to cluster_maintenance during upgrade otherwise the policy will be unchanged.
healing_in_progress_checks	6	Maximum number of attempts to check if gluster healing is still in progress.
healing_in_progress_check_delay	300	The delay in seconds between each attempt to check if gluster healing is still in progress.
wait_to_finish_healing	5	Delay in minutes to wait to finish gluster healing process after successful host upgrade.

Dependencies

No.

Example Playbook

---
- name: oVirt infra
  hosts: localhost
  connection: local
  gather_facts: false

  vars:
    engine_fqdn: ovirt-engine.example.com
    engine_user: admin@internal
    engine_password: 123456
    engine_cafile: /etc/pki/ovirt-engine/ca.pem

    cluster_name: production
    stopped_vms:
      - openshift-master-0
      - openshift-node-0
      - openshift-node-image

  roles:
    - ovirt.cluster-upgrade

License

Apache License 2.0