Mark Goddard af3b719b73 Move group_vars to an inventory directory

The group variables originally in ansible/group_vars/ were playbook
group variables, due to being adjacent to the playbooks. Typically they
provided default values for global variables in the all group, as well
as some more specific groups. This has worked fairly well, but results
in (at least) a couple of problems.

1. The default variable precedence rules mean that these playbook group
   variables have a higher precedence than inventory group variables
   (for a given group). This can make it challenging to override
   playbook group variables in the inventory in Kayobe configuration.

2. Any playbook run by Kayobe must be in the same directory as the
   playbook group variables in order to use them. Given that they
   include variables required for connectivity such as ansible_host and
   ansible_user, this is quite critical. For Kayobe custom playbooks, we
   work around this by symlinking to the group_vars directory from the
   directory containing the custom playbook. This is not an elegant
   workaround, and has assumptions about the relative paths of the
   Kayobe configuration and virtual environment in which Kayobe is
   installed.

Story: 2010280
Task: 46233

Change-Id: Ifea5c7e73f6f410f96a7398bfd349d1f631d9fc0

2022-09-08 10:11:23 +01:00

3.3 KiB

Raw Blame History

Infrastructure VMs

Kayobe can deploy infrastructure VMs to the seed-hypervisor. These can be used to provide supplementary services that do not run well within a containerised environment or are dependencies of the control plane.

Configuration

To deploy an infrastructure VM, add a new host to the the infra-vms group in the inventory:

[infra-vms]
an-example-vm

The configuration of the virtual machine should be done using host_vars. These override the group_vars defined for the infra-vms group. Most variables have sensible defaults defined, but there are a few variables which must be set.

Mandatory variables

All networks must have an interface defined, as described in configuration-network-per-host. By default the VMs are attached to the admin overcloud network. If, for example, admin_oc_net_name was set to example_net, you would need to define example_net_interface. It is possible to change the list of networks that a VM is attached to by modifying infra_vm_network_interfaces. Additional interfaces can be added by setting infra_vm_network_interfaces_extra.

List of Kayobe applied defaults to required docker_container variables. Any of these variables can be overridden with a host_var.

../../../../ansible/inventory/group_vars/all/infra-vms

Customisations

Examples of common customisations are shown below.

By default the Ansible inventory name is used as the name of the VM. This may be overridden via infra_vm_name:

# Name of the infra VM.
infra_vm_name: "the-special-one"

By default the VM has 16G of RAM. This may be changed via infra_vm_memory_mb:

# Memory in MB. Defaults to 16GB.
infra_vm_memory_mb: "{{ 8 * 1024 }}"

The default network configuration attaches infra VMs to the admin network. If this is not appropriate, modify infra_vm_network_interfaces. At a minimum the network interface name for the network should be defined.

# Network interfaces that the VM is attached to.
infra_vm_network_interfaces:
  - aio

# Mandatory: All networks must have an interface defined.
aio_interface: eth0

# By default kayobe will connect to a host via ``admin_oc_net``.
# As we have not attached this VM to this network, we must override
# ansible_host.
ansible_host: "{{ 'aio' | net_ip }}"

Configuration for all VMs can be set using extra_vars defined in $KAYOBE_CONFIG_PATH/infra-vms.yml. Note that normal Ansible precedence rules apply and the variables will override any host_vars. If you need to override the defaults, but still maintain per-host settings, use group_vars instead.

Deploying the virtual machine

Once the initial configuration has been done follow the steps in deployment-infrastructure-vms.

3.3 KiB Raw Blame History