There are several ways to implement a high available OpenStack cluster. As proven in the previous sections, HA can be achieved in each OpenStack layer by the means of many techniques including services redundancy, database replication, load balancing using HAProxy, and clustering methods such as Pacemaker and Corosync. HA setup in OpenStack should be also capable of automation. In this section, we will cover the last piece of our OpenStack: Ansible playbooks, offering an automated way of our previous HA implementation using Ansible.
As discussed in Chapter 1, OpenStack Clustering the Cloud Controller and Common Services, we have adjusted our initial deployment with three cloud controllers for running common services including databases and queuing messages. We need to describe our network ranges and our fleet of servers in the /etc/openstack_deploy/openstack_user_config.yml file. In the same file...
