Example: Installation and configuration of multiple high availability zones (Cinder)

You might have a scenario that requires two or more Cinder availability zones to satisfy the requirements from high availability, load balancing, and so on. The following steps show an example of adding Cinder availability zones to an existing IBM® Cloud Manager with OpenStack environment.

Environment requirements:
  1. You installed IBM Cloud Manager with OpenStack and have a working cloud environment that uses the Chef server.
  2. OpenStack is running with at least one Cinder availability zone.
  3. The database service port is accessible for the node on which the new Cinder service is going to be installed.
Example steps:
  1. Find a new node that does not exist in OpenStack or find an existing node that is not running the Cinder service. Configure the node so that Chef can install recipes on it.
  2. Edit the environment. Change the value of openstack.block-storage.iscsi_ip_address to the IP address of the node on which the new role will be deployed.
  3. Create a topology file and complete the information necessary for the topology file. Here is an example for the new Cinder availability zone.
    {
  "name":"CHANGEME",
  "description":"CHANGEME",
  "environment":"CHANGEME",
  "secret_file":"CHANGEME",
  "run_sequentially":false,
  "nodes": [
    {
      "fqdn":"CHANGEME",
      "password":"CHANGEME",
      "identity_file":"CHANGEME",
      "quit_on_error":true,
      "run_order_number":1,
      "runlist": [
      "role[ibm-os-block-storage-node]"
      ]
    }
  ]
}
  4. Stop the Cinder services, except for openstack-cinder-volume.
  5. A Cinder availability zone can contain one or more openstack-cinder-volume services. Change the value of storage_availability_zone in the /etc/cinder/cinder.conf file for the new installed cinder volume service. The value is the name of the availability zone. Then, restart the new openstack-cinder-volume service and all existing Cinder services.