Chapter 3. Environment Prerequisites and Limitations

Warning

When Instance HA is enabled, you cannot perform overcloud upgrade or scale-up operations. Any attempts to do this will fail. This limitation applies to minor and major upgrades. Before upgrading or scaling your overcloud, you must disable Instance HA first. For instructions, see Chapter 6, Disabling Instance HA.

To enable Instance HA, your Red Hat OpenStack Platform overcloud must meet the following requirements:

  • The environment is deployed by Red Hat OpenStack Platform director. See Director Installation and Usage for details.
  • Fencing is already manually enabled on the control plane.

  • The following package versions are installed on all nodes:

    • fence-agents-4.0.11-86.el7.x86_64 (or later)
    • pacemaker-1.1.18-11.el7.x86_64 (or later)
    • resource-agents-3.9.5-124.el7.x86_64 (or later)
  • The environment can tolerate a full outage of the Compute and the Control planes.
  • Shared storage is enabled in the environment for ephemeral and block storage. See Section 3.1, “Considerations for Shared Storage” for considerations.

  • The Message Broker (AMQP) recognizes the hostname of each Compute node as valid. To check the hostname of a Compute node, run the following command: 

    heat-admin@compute-n $ sudo crudini --get /etc/nova/nova.conf DEFAULT host
  • Each Compute node can reach the endpoint that is set in the $OS_AUTH_URL environment variable.

  • The $OS_AUTH_URL environment variable must be set to one of the following destinations:

    • The authentication services of the overcloud (which requires access to the external network), or
    • The internal authentication URL

3.1. Considerations for Shared Storage

Typically, Instance HA requires that you configure shared storage for disk images of instances. Therefore, if you attempt to use the no-shared-storage option, you might receive an InvalidSharedStorage error during evacuation, and the instances will not boot on another Compute node.

However, if all your instances are configured to boot from an OpenStack Block Storage (cinder) volume, you do not need to configure shared storage for the disk image of instances, and you can still evacuate all instances using the no-shared-storage option.

During evacuation, if your instances are configured to boot from a Block Storage volume, any evacuated instances should boot from the same volume on another Compute node. As a result, the evacuated instances immediately restart their jobs because the OS image and the application data are stored on the OpenStack Block Storage volume..

Note

The Ansible-based deployment procedure in this guide supports installation with no-shared-storage option.