openstack-ansible/doc/source/developer-docs/playbooks.rst
Jimmy McCrory e9f6acc9bc Fix 'D001 Line too long' documentation lint failures
Change-Id: If1839a4cb0fee664191491b346d7c5a8b4a30aee
2016-07-20 20:18:22 -07:00

2.1 KiB

Home OpenStack-Ansible Developer Documentation

Playbooks

The OpenStack-Ansible playbooks are stored in the playbooks directory.

There are several top-level playbooks that are run to prepare the host machines before actually deploying OpenStack and associated containers.

Running Playbooks

There is an openstack-ansible command installed by the scripts/bootstrap-ansible.sh script. This wraps the ansible-playbook command and provides the /etc/openstack_deploy/user_*.yml variable files to the playbooks.

All of the playbooks should be run within the openstack-ansible/playbooks directory

Setting up the Hosts

Run openstack-ansible setup-hosts.yml to set up the physical hosts for containers.

Setting up Infrastructure

Infrastructure pertains to utility services such as RabbitMQ, memcached, galera, and logging which are not actually OpenStack services, but that OpenStack relies on.

Run openstack-ansible setup-infrastructure.yml to install these containers.

Setting up OpenStack

Running openstack-ansible setup-openstack.yml will install the following OpenStack services:

  • Keystone
  • Swift
  • Glance
  • Cinder
  • Nova
  • Neutron
  • Heat
  • Horizon

After successful deployment, you are able to update variables in /etc/openstack_deploy/user_variables.yml.

  • Object Storage (swift)

    • The pretend_min_part_hours_passed option can now be passed to swift-ring-builder prior to performing a rebalance. This is set by the swift_pretend_min_part_hours_passed boolean variable. The default for this variable is False. However, we recommend using -e swift_pretend_min_part_hours_passed=True when running the os-swift.yml playbook to avoid resetting min_part_hours unintentionally.

      Important

      If you run this command and deploy rebalanced rings before a replication pass completes, you may introduce unavailability in your cluster.

      This should only be used for testing or fully rebalanced clusters.