You must install Ansible on the control machine, preferably in a virtual Python environment:
virtualenv pyenv
. pyenv/bin/activate
pip install ansible==2.5 netaddr
Place your PEM-formatted private key under keys/id_rsa and corresponding public key under keys/id_rsa.pub.
Ensure that private key has proper permissions (0600).
Run the (local) ansible playbook to create all needed VDI disk images:
ansible-playbook -v -i hosts.yml prepare-disk-images.yml
Verify images are created. For example:
vboxmanage showmediuminfo disk $PWD/data/postgres-n1/1.vdi
An single inventory file should be created at hosts.yml. Both vagrant and ansible will use this same inventory.
An example inventory file can be found here.
Also, group variables under group_vars must be configured. See group_vars/*.yml.example files for available options.
If we want a full Vagrant environment (of course we will also need vagrant installed), setup the machines and provision in multiple phases.
All phases, apart from the initial vagrant up, delegate their work to Ansible playbooks.
First, create some needed directories for secrets and temporary data:
mkdir -p files/secrets vagrant-data
Setup machines (networking, ansible prerequisites):
vagrant up --provision-with=shell,file
Provision in several phases:
vagrant provision --provision-with=setup-basic
vagrant provision --provision-with=setup-data-partition
vagrant provision --provision-with=setup-database
If the target machines (either virtual or physical) are already setup and networked (usually in a private network), then we can directly play the Ansible playbooks:
ansible-playbook --become --become-user root play-basic.yml
ansible-playbook --become --become-user root [-e data_partition=/dev/sdc1] play-data-partition.yml
ansible-playbook --become --become-user root play-database.yml